MLOps/Systems • Score 85
Evaluating Temporal Semantic Caching and Workflow Optimization in Agentic Plan-Execute Pipelines
arXiv:2605.20630v1 Announce Type: new
Abstract: Industrial asset operations workflows are latency-sensitive because a single user query may require coordination over sensor data, work orders, failure modes, forecasting tools, and domain-specific agents. We evaluate this problem on AssetOpsBench (AOB), an industrial agent benchmark whose plan-execute pipeline exposes repeated overhead from tool discovery, LLM planning, MCP tool execution, and final summarization. Existing LLM caching techniques such as KV-cache reuse and embedding-based semantic caching were designed for chatbot serving and break down when output validity depends on time, asset, or sensor parameters. We propose two complementary optimization layers for AOB plan-execute pipelines: a temporal semantic cache and a set of MCP workflow optimizations combining disk-backed tool-discovery caching and dependency-aware parallel step execution. MCP workflow optimizations corresponded to a 1.67x speedup and reduced median end-to-end latency by about 40.0% while the temporal-cache benchmark achieved a median of 30.6x speedup on cache hits. Beyond the speedup, our results expose a concrete failure mode of pure semantic caching for parameter-rich industrial queries, providing a critical analysis of how caching choices interact with evaluation correctness in MCP-backed agent benchmarks.
Fonte: arXiv cs.AI
RecSys • Score 85
Beyond Single Slot: Joint Optimization for Multi-Slot Guaranteed Display Advertising
arXiv:2605.21556v1 Announce Type: new
Abstract: Guaranteed display advertising is crucial for platform monetization, yet existing methods often operate under a single-slot assumption, limiting their ability to optimize allocation across multi-slot page views. In this paper, we propose a novel joint optimization framework for multi-slot GD allocation, addressing key challenges such as slot-level redundancy, contract imbalance, and exposure concentration. Our approach formulates the allocation as an offline bipartite matching problem with a contract roulette mechanism for slot exclusivity and Page View constraints for impression control, and incorporates a scalable allocation optimization algorithm for efficient large-scale deployment. Extensive online tests on the Meituan advertising platform demonstrate that our method significantly improves merchant ROI, platform revenue efficiency, and contract fulfillment robustness. Specifically, online A/B tests show a 28.99% increase in Average Revenue Per User under 70% traffic, and DID analysis further indicates improved contract stability, demonstrating the strong applicability and effectiveness of our framework in real-world advertising deployments.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
GenEvolve: Self-Evolving Image Generation Agents via Tool-Orchestrated Visual Experience Distillation
arXiv:2605.21605v1 Announce Type: new
Abstract: Open-ended image generation is no longer a simple prompt-to-image problem. High-quality generation often requires an agent to combine a model's internal generative ability with external resources. As requests become more diverse and demanding, we aim to develop a general image-generation agent that can self-evolve through trajectories and use tools more effectively across varied generation challenges. To this end, we propose GenEvolve, a self-evolving framework based on Tool-Orchestrated Visual Experience Distillation. In GenEvolve, each generation attempt is modeled as a tool-orchestrated trajectory, where the agent gathers evidence, selects references, invokes generation skills, and composes them into a prompt-reference program. Unlike existing agentic generation methods that mainly rely on image-level scalar rewards, GenEvolve compares multiple trajectories for the same request and abstracts best-worst differences into structured visual experience, provided only to a privileged teacher branch. Inspired by on-policy self-distillation, Visual Experience Distillation provides dense token-level supervision, helping the student internalize better search, knowledge activation, reference selection, and prompt construction. We further construct GenEvolve-Data and GenEvolve-Bench. Experiments on public benchmarks and GenEvolve-Bench show substantial gains over strong baselines, achieving state-of-the-art performance among current image-generation frameworks. Our website is as follows: https://ephemeral182.github.io/GenEvolve/
Fonte: arXiv cs.CV
Applications • Score 85
Quantitative coronary calcification analysis for prediction of myocardial ischemia using non-contrast CT calcium scoring
arXiv:2605.21745v1 Announce Type: new
Abstract: Non-contrast computed tomography calcium scoring (CTCS) is widely recognized as an effective tool for cardiovascular risk stratification. This study aimed to develop a novel machine learning framework for predicting myocardial ischemia from routine non-contrast CTCS scans using quantitative coronary calcium assessment. This study analyzed 1,375 patients who underwent both non-contrast CTCS and regadenoson stress cardiac positron emission tomography myocardial perfusion imaging within one year at University Hospitals Cleveland Medical Center. A total of 74 variables, including clinical variables, Agatston score, and calcium-omics features, were evaluated. Relevant features were identified using XGBoost with Shapley Additive exPlanations (SHAP). Predictive models were trained and evaluated using 5-fold cross-validation. Among 987 patients, 89 (9%) were positive for myocardial ischemia. The final model incorporated the Agatston score, eight calcium-omics features, and age. The proposed model achieved a precision of 98.9+/-3.0%, sensitivity of 79.2+/-8.4, and F1 score of 87.7+/-5.3%. The addition of calcium-omics features significantly improved predictive performance compared with models using clinical variables alone or clinical variables with the Agatston score (p<0.05). Interestingly, the number of calcified arteries, despite being the lowest-ranked feature based on SHAP analysis, showed the strongest association with myocardial ischemia in logistic regression analysis (odds ratio: 3.63, 95% confidence interval: 2.80-4.77, p<0.00001). We developed a machine learning approach for predicting myocardial ischemia using routinely acquired non-contrast CTCS scans. Calcium-omics features provided incremental predictive value beyond conventional risk factors and Agatston scoring and may support more accessible cardiovascular risk stratification.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
VBFDD-Agent for Electric Vehicle Battery Fault Detection and Diagnosis: Descriptive Text Modeling of Battery Digital Signals
arXiv:2605.20742v1 Announce Type: new
Abstract: With the rapid proliferation of electric vehicles, the safety and reliability of lithium-ion batteries have become critical concerns. Effective anomaly detection is essential for ensuring safe battery operation. However, as battery systems and operating scenarios become increasingly complex, battery fault diagnosis and maintenance require stronger cross-domain adaptability and human-AI collaboration. Traditional fault detection and diagnosis methods are usually designed for specific scenarios and predefined workflows, making them less effective in complex real-world applications.
To address the scarcity of open-source battery fault report corpora and the lack of unified maintenance knowledge representation, this study proposes a descriptive text modeling approach for battery signal reports. Monitoring signals, statistical features, anomaly records, and state assessment results are transformed into structured and readable natural language descriptions, forming a language corpus for battery health diagnosis and maintenance.
Based on this corpus, we propose VBFDD-Agent, a vehicle battery fault detection and diagnosis agent for automotive-grade battery systems. VBFDD-Agent integrates descriptive battery-state texts, historical case retrieval, local maintenance manuals, and large language model reasoning to generate structured diagnostic results and maintenance recommendations. Experiments show that the proposed framework can accurately perform anomaly monitoring based on descriptive textual representations and provide flexible, efficient, and actionable maintenance suggestions. Expert evaluation further confirms the practical value of the generated recommendations. Overall, VBFDD-Agent extends traditional battery diagnosis from label prediction to interpretable and maintenance-oriented decision support.
Fonte: arXiv cs.AI
RL • Score 85
Mahjax: A GPU-Accelerated Mahjong Simulator for Reinforcement Learning in JAX
arXiv:2605.20577v1 Announce Type: new
Abstract: Riichi Mahjong is a multi-player, imperfect-information game characterized by stochasticity and high-dimensional state spaces. These attributes present a unique combination of challenges that mirror complex real-world decision-making problems in reinforcement learning. While prior research has heavily relied on supervised learning from human play logs to pre-train the policy, algorithms capable of learning \textit{tabula rasa} (from scratch) offer greater potential for general applicability, as evidenced by the AlphaZero lineage. To facilitate such research, we introduce \textbf{Mahjax}, a fully vectorized Riichi Mahjong environment implemented in JAX to enable large-scale rollout parallelization on Graphics Processing Units (GPUs). We also provide a high-quality visualization tool to streamline debugging and interaction with trained agents. Experimental results demonstrate that Mahjax achieves throughputs of up to \textbf{2 million} and \textbf{1 million steps per second} on eight NVIDIA A100 GPUs under the no-red and red rules, respectively. Furthermore, we validate the environment's utility for reinforcement learning by showing that agents can be trained effectively to improve their rank against baseline policies.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Insights Generator: Systematic Corpus-Level Trace Diagnostics for LLM Agents
arXiv:2605.21347v2 Announce Type: new
Abstract: Diagnosing failures in LLM agents remains largely manual. Practitioners inspect a small subset of execution traces, form ad-hoc hypotheses, and iterate. This process misses patterns that only emerge across trace populations and does not scale to production corpora where individual traces span tens of thousands of tokens. We formalize the problem of corpus-level trace diagnostics. Given a corpus of execution traces, the goal is to produce grounded natural-language insights that characterize systematic behavioral patterns across trace groups, each linked to supporting evidence. We present the Insights Generator (IG), a multi-agent system that answers diagnostic questions by proposing and testing hypotheses across the trace corpus to produce an evidence-backed insights report. We evaluate IG across qualitative and objective dimensions, spanning rubric-based report assessment and downstream performance improvements achieved by implementing IG insights. Human experts using IG reports improve scaffold performance by 30.4pp over the unmodified baseline scaffold, and coding agents leveraging IG-derived insights show consistent and stable gains. Across benchmarks, IG's scout-investigator architecture produces findings comparable in detection coverage to competing approaches, while domain experts rated IG reports as leading depth and evidence quality.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
From Parameters to Data: A Task-Parameter-Guided Fine-Tuning Pipeline for Efficient LLM Alignment
arXiv:2605.21558v1 Announce Type: new
Abstract: Adapting Large Language Models (LLMs) to specialized domains typically incurs high data and computational overhead. While prior efficiency efforts have largely treated data selection and parameter-efficient fine-tuning as isolated processes, our empirical analysis suggests they may be intrinsically coupled. We posit the Strong Map Hypothesis: a sparse subset of attention heads plays a dominant role in task-specific adaptation, acting as keys that unlock specific data patterns. Building on this observation, we propose From Parameters to Data (P2D), a unified framework that leverages these task-sensitive attention heads as a dual compass for both sample mining and structural pruning. To rigorously quantify the total pipeline cost, we introduce the Alignment Efficiency Ratio (AER) metric for both selection latency and training time. Mechanistically, P2D identifies critical heads via a lightweight proxy and uses them as a functional filter to curate high-affinity data, establishing a synergistic pipeline. Empirically, by updating merely 10% of attention heads on 10% of the data, P2D achieves an 8.3 pp performance gain over strong baselines and delivers a 7.0x end-to-end time speedup. These results validate that precise parameter-data synchronization eliminates redundancy, offering a new paradigm for efficient alignment.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
TONIC: Token-Centric Semantic Communication for Task-Oriented Wireless Systems
arXiv:2605.21553v1 Announce Type: new
Abstract: Tokens are becoming the basic units through which foundation models represent and process information for understanding and inference. However, traditional wireless communication, centered on bit-level fidelity, faces a mismatch between what is transmitted reliably and what downstream models actually consume. This mismatch calls for a communication design that directly accounts for token-level task relevance and downstream model requirements, rather than treating all transmitted bits as equally important. In this paper, we propose TONIC, a token-centric semantic communication framework for task-oriented wireless systems. The transmitter converts each source sample into a sequence of tokens, estimates token-level task relevance, and allocates protection through utility-aware unequal error protection under a fixed channel-use budget. At the receiver, token-level confidence is used to gate unreliable decisions, turning harmful substitutions into recoverable erasures before a Transformer-based completion model restores the masked tokens for final task inference. Our framework combines transmitter-side semantic-aware protection with receiver-side confidence-aware gating in a modular and interpretable architecture, rather than relying solely on fully black-box end-to-end learning. We further establish a utility-aware Bayes-risk interpretation for the receiver-side gating rule and study its interaction with unequal protection and completion. Experimental results on image classification show that TONIC consistently outperforms separation-based schemes, the pixel-domain DeepJSCC baseline, and token-domain baselines under matched communication budgets over AWGN, Rayleigh, and Rician channels.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Embedding-Based Federated Learning with Runtime Governance for Iron Deficiency Prediction
arXiv:2605.21563v1 Announce Type: new
Abstract: Recent reviews find that the vast majority of published healthcare federated learning (FL) studies never reach real-world deployment. We developed an embedding-based FL pipeline for iron deficiency prediction from routine full blood count (FBC) data and deployed it across real institutional environments at Amsterdam University Medical Centre (AUMC) and NHS Blood and Transplant (NHSBT), two clinical environments that differ markedly in iron deficiency prevalence, ferritin distribution, and subject populations. A frozen domain-specific haematology foundation model, DeepCBC, performs site-local representation extraction, restricting federated training to a compact downstream classifier and substantially reducing recurrent communication relative to full-encoder federation. The two clinical datasets are structurally not independent and identically distributed (non-IID), with heterogeneity arising from distinct population differences rather than sampling artefacts. Runtime governance is enforced by FLA$^3$, a healthcare-oriented FL platform providing study-scoped execution, policy-based authorisation, and signed audit logging. Standard sample-size-weighted aggregation (FedAvg) reduced the area under the receiver operating characteristic curve (ROC-AUC) at both sites relative to local-only training, as the global update was biased towards the larger AUMC distribution. FedMAP, a personalised aggregation method, raised ROC-AUC from 0.9470 to 0.9594 at AUMC and from 0.8558 to 0.8671 at NHSBT relative to local-only training, achieving the highest macro ROC-AUC of 0.9133 and the best macro balanced accuracy overall. These results support personalised aggregation in clinical federations where client sample size and task relevance diverge substantially.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Models Can Model, But Can't Bind: Structured Grounding in Text-to-Optimization
arXiv:2605.21751v1 Announce Type: new
Abstract: Text-to-optimization requires two separable capabilities: modeling -- choosing the right optimization structure -- and binding -- grounding every coefficient, index, and parameter in the concrete problem data. We study this via Text2Opt-Bench, a scalable benchmark of solver-verified optimization problems spanning 12 categories, from textbook linear programs to stochastic and multi-objective formulations with up to thousands of variables. Across 10+ models, we find that accuracy collapses as instance data grows, even when the formulation itself is simple. We call this the effective binding limit. We address this via a simple inference-time approach, BIND, which externalizes numeric data to structured files so the model binds data programmatically rather than transcribing from the prompt. BIND improves GPT-5-Nano from 59.1% to 82.4% accuracy, matching pass@5 (82.0%) at lower token cost than pass@1, and GPT-5 from 86.2% to 95.8%. Furthermore, we validate our hypothesis by finetuning a model exclusively on binding and show that it outperforms end-to-end SFT and RL across three structurally distinct optimization categories, with a 1.5B binding specialist alone matching a 7B end-to-end baseline.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Reliable Wireless Indoor Localization via Cross-Validated Prediction-Powered Calibration
arXiv:2507.20268v3 Announce Type: replace-cross
Abstract: Wireless indoor localization using predictive models with received signal strength information (RSSI) requires proper calibration for reliable position estimates. One remedy is to employ synthetic labels produced by a (generally different) predictive model. But fine-tuning an additional predictor, as well as estimating residual bias of the synthetic labels, demands additional data, aggravating calibration data scarcity in wireless environments. This letter proposes an approach that efficiently uses limited calibration data to simultaneously fine-tune a predictor and estimate the bias of synthetic labels, yielding prediction sets with rigorous coverage guarantees. Experiments on a fingerprinting dataset validate the effectiveness of the proposed method.
Fonte: arXiv stat.ML
MLOps/Systems • Score 85
From Sequential Nodes to GPU Batches: Parallel Branch and Bound for Optimal $k$-Sparse GLMs
arXiv:2605.22188v1 Announce Type: cross
Abstract: GPUs have significantly accelerated first-order methods for large-scale optimization, especially in continuous optimization. However, this success has not transferred cleanly to problems with discrete variables, combinatorial structure, and nonlinear objectives, such as certifying optimal solutions for cardinality-constrained generalized linear models. Major challenges include the sequential processing of heterogeneous nodes in branch and bound (BnB) and frequent data movement between the CPU and GPU. We propose a simple, generic, and modular CPU--GPU framework that processes multiple BnB nodes in batches on GPUs. The framework is built around a small set of GPU-efficient routines and uses padding together with lightweight custom kernels to handle irregular node data structures. Experiments show one to two orders of magnitude speedups and zero optimality gap on challenging instances. The framework can also be extended to collect the entire Rashomon set, enabling downstream statistical analysis such as variable-importance analysis and model selection under secondary user-specific measures (e.g., AUC in classification).
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
AutoMCU: Feasibility-First MCU Neural Network Customization via LLM-based Multi-Agent Systems
arXiv:2605.21560v1 Announce Type: new
Abstract: Deploying neural networks on microcontroller units (MCUs) is critical for edge intelligence but remains challenging due to tight memory, storage, and computation constraints. Existing approaches, such as model compression and hardware-aware neural architecture search (HW-NAS), often depend on proxy metrics, incur high search cost, and do not fully bridge the gap between architecture design and verified deployment. This paper presents AutoMCU, a feasibility-first large language model (LLM)-based multi-agent system for automated neural network customization under MCU constraints. Given natural-language task requirements and hardware specifications, AutoMCU iteratively generates structured architecture candidates, filters infeasible designs through vendor toolchain feedback before training, evaluates feasible models under a controlled protocol, and verifies deployability through backend-grounded deployment analysis. AutoMCU includes two key mechanisms: 1) hardware-in-the-loop architecture generation for early elimination of undeployable candidates under RAM and Flash constraints, and 2) state-isolated multi-agent scheduling for stable coordination of proposal, training, evaluation, and deployment stages. Experiments on CIFAR-10 and CIFAR-100 under strict MCU constraints show that AutoMCU achieves competitive accuracy while reducing customization time to about 1--2 hours, compared with hundreds of GPU hours for representative MCU-oriented HW-NAS baselines. Comparisons with ColabNAS and the LLM-based NAS method GENIUS on NAS-Bench-201 further demonstrate the effectiveness and stability of AutoMCU. Real-device deployments on multiple STM32 microcontrollers validate its practical applicability to MCU-scale edge intelligence.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
SO-Mamba: State-Ownership Mamba for Unrolled MRI Reconstruction
arXiv:2605.22031v1 Announce Type: new
Abstract: Accelerated MRI reconstruction requires recovering missing details while preserving anatomically coherent structures across large spatial regions. State-space models such as Mamba provide efficient long-range modeling, making them attractive learned regularizers for unrolled reconstruction. However, in a data-consistency-coupled unrolled solver, different stages operate on different reconstruction iterates, where the resident carrier should preserve coherent reconstruction content across stages while stage-dependent non-resident evidence is tied to the current update. Treating these roles uniformly can place persistent resident-carrier evidence and update-dependent non-resident evidence into the same recurrent content route. We therefore propose SO-Mamba, a state-ownership Mamba regularizer that assigns reconstruction evidence within each Mamba stage to recurrent residency, state-interface access, and non-state output correction. SO-Mamba implements this ownership rule with a State-Ownership Router (SOR), which constructs a resident carrier for recurrent content and routes non-resident evidence to affine modulation of the B/C state interfaces and an output correction outlet. The resident carrier supplies the Mamba content route, while the non-resident evidence stream adapts the state interfaces and contributes through the output outlet without entering the recurrent content route. We further introduce a two-level outer-band leakage diagnostic that separates hidden-state storage from readout expression by measuring outer-band energy in the selective-scan state trajectory and the post-scan Mamba readout. Experiments on five public MRI reconstruction benchmarks spanning diverse anatomies, sampling patterns, and coil configurations show that SO-Mamba consistently improves over CNN-, Transformer-, and Mamba-based baselines with competitive computational efficiency.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
Governance by Design: Architecting Agentic AI for Organizational Learning and Scalable Autonomy
arXiv:2605.20210v1 Announce Type: cross
Abstract: Agentic AI systems - systems that can pursue goals through multi-step planning and tool-mediated action with limited direct supervision - are moving from experimental prototypes to enterprise deployments. This transition introduces tensions in implementation, scaling, and governance: organizations seek scalable autonomy for knowledge and coordination work, yet must preserve accountability, safety, cost control, and responsibility as systems initiate actions, access enterprise data, and evolve through iterative updates. Building on an in-depth qualitative case of a large IT services company's 2025 development and staged rollout of an agentic system integrated with enterprise tools; we show that governance is implemented through concrete architectural and working arrangements that determine what the system is allowed to do, which tools and data it can use, how memory is handled, and how performance improvements are introduced over time. We then distill seven lessons that explain how to build effective governance into agentic AI during operationalization and scaling.
Fonte: arXiv cs.AI
NLP/LLMs • Score 90
AutoRPA: Efficient GUI Automation through LLM-Driven Code Synthesis from Interactions
arXiv:2605.21082v1 Announce Type: new
Abstract: Large Language Model (LLM) based agents have demonstrated proficiency in multi-step interactions with graphical user interfaces (GUIs). While most research focuses on improving single-task performance, practical scenarios often involve repetitive GUI tasks for which invoking LLM reasoning repeatedly, i.e., the ReAct paradigm, is inefficient. Prior to LLMs, traditional Robotic Process Automation (RPA) offers runtime efficiency but demands significant manual effort to develop and maintain. To bridge this gap, we propose AutoRPA, a framework that automatically distills the decision logic of ReAct-style agents into robust RPA functions. AutoRPA introduces two core innovations: (1) A translator-builder pipeline, where a translator agent converts hard-coded ReAct actions into soft-coded procedures, and a builder agent synthesizes robust RPA functions via retrieval-augmented generation over multiple trajectories; (2) A hybrid repair strategy during code verification, combining RPA execution with ReAct-based fallback for iterative refinement. Experiments across multiple GUI environments demonstrate that RPA functions generated by AutoRPA successfully solve similar tasks while reducing token usage by 82% to 96%, significantly improving runtime efficiency and reusability.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
SpecHop: Continuous Speculation for Accelerating Multi-Hop Retrieval Agents
arXiv:2605.21965v1 Announce Type: new
Abstract: Large language models increasingly use external tools such as web search and document retrieval to solve information-intensive tasks. However, multi-hop tool use in complex tasks introduces substantial latency, since the model must repeatedly wait for tool observations before continuing. We study how to accelerate such trajectories without changing the final trajectory the model would have taken without acceleration, assuming access to faster but less reliable speculator tools. We develop a theoretical framework for lossless speculation in multi-hop tool-use settings, characterizing the optimal achievable latency gain. We propose SpecHop, a continuous speculation framework that maintains multiple speculative threads, verifies predicted observations asynchronously as target tool outputs arrive, commits correct branches, and rolls back incorrect ones. This preserves accuracy while reducing wall-clock latency. We show that SpecHop can approach oracle latency gains with enough active threads. Empirically, on retrieval-augmented multi-hop tasks, SpecHop closely matches theoretical predictions and reduces latency by up to 40\% in some settings. Code: https://github.com/mehrdadsaberi/spechop
Fonte: arXiv cs.CL
MLOps/Systems • Score 85
ORBIS: Output-Guided Token Reduction with Distribution-Aware Matching for Video Diffusion Acceleration
arXiv:2605.22015v1 Announce Type: new
Abstract: Diffusion Transformer (DiT) has emerged as a powerful model architecture for generating high-quality images and videos. In the case of video DiT, 3D Spatio-Temporal Attention increases token length in proportion to the number of frames, sharply increasing computational cost. Token reduction methods mitigate this cost by exploiting spatial redundancy, but existing approaches rely on inaccurate similarity estimates and lightweight matching algorithms, resulting in poor matching quality and only marginal acceleration.
To overcome these limitations, we propose ORBIS, an SW-HW co-designed accelerator for video DiT. ORBIS leverages the output activation from the previous timestep to obtain more accurate inter-token similarity, substantially improving matching quality and enabling a higher token reduction ratio. We further introduce a Distribution-Aware Token Matching (DATM) algorithm that captures global token distribution and explicitly minimizes token-pair loss for additional gains. To fully hide DATM latency, we design specialized, deeply pipelined hardware and minimize its hardware cost through quantization, occupying only 2.4% of total area with negligible accuracy loss. Extensive experiments show that ORBIS achieves about 2x higher token reduction ratio than the state-of-the-art approach, AsymRnR, while delivering up to 4.5x speedup and 79.3% energy reduction compared to an NVIDIA A100 GPU.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Hierarchical Variational Policies for Reward-Guided Diffusion
arXiv:2605.21661v1 Announce Type: new
Abstract: Adapting pretrained diffusion models to downstream objectives such as inverse problems often requires expensive test-time guidance or optimization. We propose a principled framework for generating high-quality reward-aligned samples at substantially reduced inference cost. Our approach formulates test-time adaptation as a hierarchical variational model, where control is amortized into a lightweight yet expressive stochastic policy. This formulation naturally supports few-step diffusion sampling: large step sizes enable fast inference, while the learned policy maintains sample quality by providing structured per-step control. The resulting fully amortized sampler achieves a strong quality--speed tradeoff, matching or exceeding recent test-time scaling baselines while requiring significantly less compute. For example, on 4x super-resolution, our method achieves better perceptual quality with more than 5x faster inference compared to the best-performing baseline. We further extend our approach to a semi-amortized regime that combines cheap amortized proposals with limited test-time optimization, achieving state-of-the-art perceptual quality across several challenging inverse problems.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Towards Resilient and Autonomous Networks: A BlueSky Vision on AI-Native 6G
arXiv:2605.21395v1 Announce Type: new
Abstract: The proliferation of emerging applications, such as autonomous driving and immersive experiences, demands cellular networks that are not only faster, but fundamentally more resilient and autonomous. This paper presents a BlueSky vision on how Artificial Intelligence will be natively integrated into 6G, shifting the paradigm from \underline{Network for AI} to \underline{AI for Network}. We envision that, unlike 5G's reliance on scattered, ad-hoc models each trained for a single task, native AI in the 6G era will be anchored by a foundation model and and orchestrated via collaborative multi-agent systems, framing network management as a unified, multi-modal, multi-task optimization problem. Built on this vision, we outline two transformative directions. The first focuses on developing a 6G foundation model as a unified backbone, with task-specific knowledge distilled into compact models suited for diverse edge deployments. The second advances multi-agent systems designed to autonomously diagnose, maintain, and recover networks with minimal human intervention. These directions chart a roadmap for 6G to evolve into an intelligent, self-sustaining communication infrastructure.
Fonte: arXiv cs.AI
Vision • Score 85
UniVL: Unified Vision-Language Embedding for Spatially Grounded Contextual Image Generation
arXiv:2605.21611v1 Announce Type: new
Abstract: We introduce spatially grounded contextual image generation, a controllable image generation task that reframes the conditioning paradigm. Instead of supplying a reference image and a global text prompt through two separate encoders, one for vision and one for language, UniVL is trained to bind semantics to spatial locations directly from a single unified visual input, where the textual instruction is rendered onto the spatial mask. This removes the need for a standalone text encoder at inference time. The resulting model supports contextual image generation by following user-specified instructions about what should appear where, while substantially reducing computation.
To address this task, we propose a framework in which the UniVL encoder, adapted from an optical-character-recognition-pretrained backbone, reads the unified condition optically and produces a UniVL embedding, fVIL, that fuses visual and semantic intent with spatial locations in a single token sequence. A two-stage pipeline first aligns UniVL with the VAE embedding space and then conditions a pretrained diffusion backbone entirely on UniVL embeddings, eliminating the standalone text encoder, such as T5. Although this reframing uses a deliberately minimal text interface, it yields strong empirical gains. On UniVL-ImgGen, a benchmark of 477K mask-annotated images that we construct for training and evaluation, UniVL improves image quality over text-prompted baselines, reducing FID from 14 to 11 and increasing PSNR from 16 to 20. It also eliminates the text encoder entirely, reducing inference TFLOPs by up to 52% and runtime by up to 44%. Additional ablation studies validate the contributions of the proposed components, paving the way for efficient, spatially grounded image generation with a unified conditioning paradigm.
Fonte: arXiv cs.CV
Vision • Score 85
Improving 3D Labeling in Self-Driving by Inferring Vehicle Information using Vision Language Models
arXiv:2605.21747v1 Announce Type: new
Abstract: We present an approach to improve 3D vehicle labeling in self-driving applications through zero-shot inference of vehicle information, leveraging Vehicle Make and Model Recognition (VMMR) methods. The proposed approach utilizes a Vision Language Model (VLM) to both infer a vehicle's make, model, and generation from image crops, and output accurate 3D bounding box dimensions to seed manual labeling. We evaluate the impact of iterative prompt engineering and the choice of different VLMs on both vehicle bounding box inference and make/model/generation recognition. When compared to strong baselines, the proposed approach not only shows high accuracy, but also excels in mitigating specific failure modes where VLMs provide better dimensions than initial lidar-aided human annotated labels (e.g., in cases of significant vehicle occlusion). Experiments on both public and proprietary data strongly suggest that our conclusions are generalizable across different labelers and datasets. The results demonstrate that integrating VLMs into the labeling process can reduce manual labeling time while increasing label quality.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
FlyRoute: Self-Evolving Agent Profiling via Data Flywheel for Adaptive Task Routing
arXiv:2605.22057v1 Announce Type: new
Abstract: Enterprise routers assign queries to expert agents, yet deployed profiles stay static while agents evolve (prompts, tools, models), and developers rarely keep descriptions or exemplars current. We present FlyRoute, a self-evolving profiling framework that grows capability evidence from real traffic: dispatch candidates, quality-gate successful pairs into each agent's success store, periodically distill evidence into learned capability descriptions, and inject those descriptions together with BM25-retrieved successes into an LLM router. To make this flywheel data-efficient, FlyRoute introduces a targeted exploration policy that combines profile uncertainty, BM25 relevance, and lexical novelty, prioritizing under-profiled agents only for plausible queries and avoiding redundant evidence collection. In experiments on our proprietary enterprise developer-support dataset of real routed queries, FlyRoute improves a same-backbone zero-shot LLM router from 72.57% to 78.04% with only five seed queries per agent, showing that profile retrieval already strengthens cold-start routing. After streaming 7,211 labeled training queries through the flywheel, accuracy rises to 89.83% (+17.26pp over zero-shot; +11.79pp over cold start), with consistent gains across four expert domains under standard routing accuracy on single-gold test queries.
Fonte: arXiv cs.CL
Theory/Optimization • Score 85
TBP-mHC: full expressivity for manifold-constrained hyper connections through transportation polytopes
arXiv:2605.21724v1 Announce Type: new
Abstract: Hyper-Connections (HC) improve residual networks by introducing learnable mixing across multiple residual streams, but unconstrained mixing leads to training instability. Manifold-Constrained Hyper-Connections (mHC) address this by enforcing approximate double stochasticity via Sinkhorn normalization, while mHC-lite ensures exact constraints through convex combinations of permutation matrices at the cost of factorial complexity. KromHC reduces this cost using Kronecker-product parameterizations, but restricts the mixing matrices to a structured submanifold of the Birkhoff polytope .
We propose Transportation Birkhoff Polytope (TBP) parameterizations and their Recursive variants (RTBP), which construct exactly doubly stochastic mixing matrices with $(n-1)^2$ degrees of freedom. Our approach avoids iterative normalization and combinatorial explosion while preserving full expressivity of the Birkhoff polytope. Empirical results on language model pre-training' demonstrate competitive performance with improved stability and scalability.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Broadening Access to Transportation Safety Data with Generative AI: A Schema-Grounded Framework for Spatial Natural Language Queries
arXiv:2605.21712v1 Announce Type: new
Abstract: Transportation safety analysis requires integrating crash records, roadway attributes, and geospatial data through GIS-based workflows, but access remains uneven across agencies and community stakeholders. Technical prerequisites create a gap between analytical tools central to safety planning and the practitioners able to use them. Local agencies, school committees, and residents may have safety concerns but limited capacity to retrieve, filter, map, and analyze relevant data. Generative AI offers a way to narrow this divide, but its public-sector use raises questions about reliability, reproducibility, and governance. This paper presents a schema-grounded natural language interface for transportation safety analysis, using a large language model (LLM) to interpret user intent while preserving deterministic, reviewable execution against an authoritative database. User queries are translated into structured semantic frames, validated by a rule-based layer, compiled into a typed directed acyclic graph of spatial operations, and executed against a PostGIS database. This bounded design separates language interpretation from deterministic execution, keeping results reproducible and schema-grounded while removing access barriers. The framework is evaluated using a statewide Massachusetts transportation safety database integrating crash records, roadway attributes, and geospatial layers including schools, bus stops, crosswalks, and municipal boundaries. All queries executed successfully; the validation layer corrects errors in 29% of evaluation queries, reflecting the gap between flexible natural language and strict schema-grounded requirements. The results suggest that combining natural language accessibility with deterministic execution is a practical direction for broadening access to transportation safety data, with implications for trustworthy AI in public-sector planning.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
DualOptim+: Bridging Shared and Decoupled Optimizer States for Better Machine Unlearning in Large Language Models
arXiv:2605.21539v1 Announce Type: new
Abstract: We propose DualOptim+, a novel optimization framework for improving machine unlearning in large language models. It introduces a base state to capture common representations shared by forgetting and retaining objectives and delta states to preserve objective-specific residuals. This architecture allows the optimizer to adaptively bridge shared and decoupled states based on the directional conflict between forgetting and retaining gradients. We further introduce DualOptim+ 8bit, a quantized variant that reduces memory overhead without compromising performance. Extensive experiments across fictitious and real-world unlearning, safety alignment, and multi-task learning tasks demonstrate that DualOptim+ consistently achieves a superior trade-off between different objectives. Codes are available at https://github.com/CityU-MLO/DualOptimPlus.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Tool-Augmented Agent for Closed-loop Optimization,Simulation,and Modeling Orchestration
arXiv:2605.20190v1 Announce Type: new
Abstract: Iterative industrial design-simulation optimization is bottlenecked by the CAD-CAE semantic gap: translating simulation feedback into valid geometric edits under diverse, coupled constraints. To fill this gap, we propose COSMO-Agent (Closed-loop Optimization, Simulation, and Modeling Orchestration), a tool-augmented reinforcement learning (RL) framework that teaches LLMs to complete the closed-loop CAD-CAE process. Specifically, we cast CAD generation, CAE solving, result parsing, and geometry revision as an interactive RL environment, where an LLM learns to orchestrate external tools and revise parametric geometries until constraints are satisfied. To make this learning stable and industrially usable, we design a multi-constraint reward that jointly encourages feasibility, toolchain robustness, and structured output validity. In addition, we contribute an industry-aligned dataset that covers 25 component categories with executable CAD-CAE tasks to support realistic training and evaluation. Experiments show that COSMO-Agent training substantially improves small open-source LLMs for constraint-driven design, exceeding large open-source and strong closed-source models in feasibility, efficiency, and stability.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Declarative Data Services: Structured Agentic Discovery for Composing Data Systems
arXiv:2605.20690v1 Announce Type: new
Abstract: Agentic discovery has shown that LLM-driven search can find novel algorithms, designs, and code under benchmark conditions. Translating the paradigm to multi-system data backends surfaces a harder problem: the search space is heterogeneous, the verifier is whether a deployed stack actually runs, and composition knowledge is unevenly captured in pretraining. Unbounded agentic discovery, a coding agent iterating on failure-log feedback, fails to converge consistently on a working stack even when iteration and explicit composition knowledge are added. We propose Declarative Data Services (DDS), an architecture for structured agentic discovery of data-system compositions from declarative user intent. The framework owns four typed contracts at successive layers (intent, operator DAG, per-system skills, runtime attribution) that decompose the global search into bounded sub-searches; sub-agents search each typed space, while the framework provides the channels by which knowledge flows forward as inline skill citations and errors route backward as typed signals. As a proof of life on a trading-backend workload, DDS converges where unbounded discovery does not; runtime failures become skill patches that the next deployment cites inline. We position this as an early prototype reporting lessons from real-world data-system composition.
Fonte: arXiv cs.AI
Vision • Score 85
Lens: Rethinking Training Efficiency for Foundational Text-to-Image Models
arXiv:2605.21573v1 Announce Type: new
Abstract: We introduce Lens, a 3.8B-parameter T2I model that achieves performance competitive with, and in several cases surpassing, state-of-the-art models with more than 6B parameters across various benchmarks, while requiring significantly less training compute. For example, Lens requires only about 19.3% of the training compute used by Z-Image. The training efficiency of Lens stems from two key strategies beyond its compact model size. First, we maximize data information density per training batch by (i) training on Lens-800M, a dataset of 800M densely captioned image-text pairs whose captions are generated by GPT-4.1 and contain approximately 109 words on average, providing richer semantic supervision than conventional short captions, and (ii) constructing each batch from images with multiple resolutions and diverse aspect ratios, thereby enlarging the effective visual coverage of each optimization step. Second, we improve convergence speed through careful architectural choices, including adopting a semantic VAE that provides better latent representations and employing a strong language encoder that accelerates optimization while enabling multilingual generalization from English-only training data. After pre-training, we apply RL with taxonomy-driven prompts (Lens-RL-8K) and structured reward rubrics to suppress artifacts and improve visual quality, a reasoner module with training-free system prompt search to better align user requests with the model, and distillation-based acceleration for 4-step inference. Through efficient training and systematic optimization, Lens generalizes to arbitrary aspect ratios from 1:2 to 2:1 and resolutions up to 1440^2, and supports prompts in several commonly used languages. Thanks to its compact size, Lens generates a 1024^2 image in 3.15 seconds on a single NVIDIA H100 GPU, while its distilled turbo version performs 4-step generation in 0.84 seconds.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
AgentCo-op: Retrieval-Based Synthesis of Interoperable Multi-Agent Workflows
arXiv:2605.20425v1 Announce Type: new
Abstract: Designing multi-agent workflows is especially difficult in open-ended scientific settings where tasks lack curated training sets, reliable scalar evaluation metrics, and standardized interfaces between existing tools and agents. We propose AgentCo-op, a retrieval-based synthesis framework that composes reusable skills, tools, and external agents into executable workflows through typed artifact handoffs, then applies bounded self-guided local repair to implicated components when execution evidence indicates failure. In two open-world genomics case studies, AgentCo-op composes independently developed scientific agents and external tool repositories into auditable workflows without redesigning them or running global topology search. It coordinates specialized agents for spatial transcriptomics and gene-set interpretation to enable collaborative discovery from spatial transcriptomics data, and builds a parallel workflow for cross-modality marker analysis on single-cell multiome data. AgentCo-op can also import a searched workflow as a structural prior and improve it by grounding nodes with retrieved components and applying local repair, showing that synthesis and search are complementary. On six coding, math, and question-answering benchmarks, AgentCo-op achieves the best result on four benchmarks and the best average score under a unified backbone setting, while consistently reducing per-task cost relative to multi-agent baselines. Together, these results suggest that retrieval-based synthesis can extend automated agentic workflow design beyond benchmark-optimized agent graphs to open-world workflows built from existing agents, tools, and typed artifacts.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
PALS: Power-Aware LLM Serving for Mixture-of-Experts Models
arXiv:2605.21427v1 Announce Type: new
Abstract: Large language model (LLM) inference has become a dominant workload in modern data centers, driving significant GPU utilization and energy consumption. While prior systems optimize throughput and latency by batching, scheduling, and parallelism, they largely treat GPU power as a static constraint rather than a controllable resource. In this paper, we present a power-aware runtime for LLM serving, PALS, that treats GPU power caps as a first-class control knob and jointly optimizes them with software parameters such as batch size. The system combines lightweight offline power-performance models with a feedback-driven controller to select configurations that satisfy throughput targets while maximizing energy efficiency. We implement PALS within an existing LLM serving framework, vLLM, demonstrating that it requires no model retraining or API changes. Across multi-GPU systems and both dense and mixture-of-experts (MoE) models, PALS improves energy efficiency by up to 26.3%, reduces QoS violations by 4x to 7x under power constraints, and tracks dynamic power budgets. These results highlight the potential of integrating power control directly into LLM inference runtimes, enabling energy-proportional and grid-interactive AI systems.
Fonte: arXiv cs.AI
Vision • Score 85
MAVEN: A Multi-stage Agentic Annotation Pipeline for Video Reasoning Tasks
arXiv:2605.21917v1 Announce Type: new
Abstract: Training Vision Language Models (VLMs) for video event reasoning requires high-quality structured annotations capturing not only what happened, but when, where, why, and with what consequence, at a scale manual labelling cannot support. We present MAVEN (Multi-stage Agentic Video Event aNnotation), a multi-stage agentic pipeline that turns raw videos into multi-task training data with Chain-of-Thought (CoT) reasoning traces, organized around a designated Event of Focus. At its core, MAVEN synthesizes a Multi-Scale Spatio-Temporal Event Description (MSTED) from three complementary caption levels; this explicit intermediate serves as the sole input to downstream Q&A generation across multiple task formats. Crucially, MAVEN supports agent-driven domain adaptation: given a new video dataset and target question examples, the agent redesigns all prompts top-down without manual re-engineering. A hierarchical refinement loop further classifies annotation errors against a taxonomy, traces root causes to the originating pipeline stage, and applies targeted edits that rewrite prompts or modify the pipeline structure itself, iteratively improving data quality. We apply MAVEN to label over 5,300 traffic videos and fine-tune Cosmos-Reason2-8B on the resulting data. On a private CCTV evaluation set, fine-tuning surpasses both Gemini 2.5 Pro and 3.1 Flash, including a $+38.8$-point gain in MCQ accuracy over zero-shot. On AccidentBench, CCTV-only training lifts Cosmos-Reason2 by $+10.7$ MCQ points and matches Gemini 2.5 Pro despite seeing no dashcam videos; adding agent-adapted dashcam annotations narrows the gap to Gemini 3.1 Flash, and RL post-training pushes overall performance past both Gemini baselines. Qualitative results on warehouse surveillance and public safety videos further show the agentic workflow readily adapts the pipeline to new domains.
Fonte: arXiv cs.CV
Vision • Score 85
Universal CT Representations from Anatomy to Disease Phenotype through Agglomerative Pretraining
arXiv:2605.21906v1 Announce Type: new
Abstract: Computed tomography (CT) is a central to three-dimensional medical imaging, yet CT-based artificial intelligence remains fragmented across task-specific models for segmentation, classification, registration, and report analysis. Here we present FlexiCT, a family of CT foundation models trained by agglomerative continual pretraining on 266,227 CT volumes from 56 publicly available datasets, forming a large-scale public resource for CT representation learning. FlexiCT uses agglomerative pretraining across three stages: two-dimensional axial pretraining, three-dimensional anatomical pretraining and report-guided semantic alignment. This training strategy supports slice-level, volume-level and vision-language analysis. Across five downstream task families (segmentation, classification, registration, vision-language understanding and clinical retrieval), FlexiCT matches or exceeds prior task-specific approaches on multiple benchmarks. Its embeddings further organize CT scans along gradients associated with various tumor stages, suggesting that CT foundation models can capture imaging features relevant to disease phenotype characterization. Code is available at https://github.com/ricklisz/FlexiCT
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
Aerodynamic force reconstruction using physics-informed Gaussian processes
arXiv:2605.22111v1 Announce Type: cross
Abstract: Accurate modeling of aerodynamic loads is essential for understanding and predicting the responses of complex structural systems. However, these models often rely on simplifications of the true physical forces, introducing assumptions that can limit their accuracy. Validating such models becomes particularly challenging in the presence of noisy or incomplete data. To address this, we introduce a probabilistic physics-informed machine learning approach designed to reconstruct the underlying aerodynamic loads from noisy measurements of structural dynamic responses. The model avoids overfitting, eliminates the need for regularization schemes, and allows for the use of heterogeneous and multi-fidelity data during the training process. The efficacy of the approach is demonstrated through the reconstruction of aerodynamic loads on the Great Belt East Bridge, simulated under a linear unsteady assumption. Results show a strong agreement between true and predicted loads, particularly related to root mean squared errors, magnitude, phase angle and peak values of the signals. The method for load reconstructing holds broad applicability, such as modeling validation, future load estimation, and structural damage prognosis.
Fonte: arXiv stat.ML
MLOps/Systems • Score 85
AiraXiv: An AI-Driven Open-Access Platform for Human and AI Scientists
arXiv:2605.21481v1 Announce Type: new
Abstract: Recent advances in artificial intelligence (AI) have accelerated the growth of both human-authored and AI-generated research outputs, placing increasing strain on traditional academic publishing systems and challenging the scalability of conference- and journal-centered paradigms amid rising submission volumes, reviewer workload, and venue size. To address these challenges, we explore an AI-era publishing paradigm in which both human and AI scientists participate as authors and readers, and papers evolve through continuous, feedback-driven iteration. We propose AiraXiv, an AI-driven open-access platform built on open preprints, AI-augmented analysis and review, and reader feedback. AiraXiv supports human scientists through an interactive UI and AI scientists through Model Context Protocol (MCP)-based interactions. We validate AiraXiv through real-world deployments, including serving as the submission platform for ICAIS 2025, demonstrating its potential as a fast, inclusive, and scalable research infrastructure for the AI era. AiraXiv is publicly available at https://airaxiv.com.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
From Automated to Autonomous: Hierarchical Agent-native Network Architecture (HANA)
arXiv:2605.20608v1 Announce Type: new
Abstract: Realizing Level 4/5 Autonomous Networks (AN) demands a shift from static automation to agent-native intelligence. Current operations, reliant on rigid scripts, lack the cognitive agency to handle off-nominal conditions. To address this, this letter proposes a hierarchical multi-agent reference architecture enabling high-level autonomy. The framework features a Dual-Driven Orchestrator that coordinates specialized Executive Agents, supported by a shared Public Memory for unified domain knowledge. A key innovation is the integration of agent self-awareness, which empowers the system to harmonize deliberative strategic governance with reflexive fault recovery. We instantiate and validate this architecture within a 5G Core environment. Case studies demonstrate that the system sustains critical throughput under congestion and reduces Mean Time to Repair (MTTR) by 86%, confirming its efficacy in unifying strategic planning with operational resilience.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Governance by Construction for Generalist Agents
arXiv:2605.20874v1 Announce Type: new
Abstract: Enterprise agents are increasingly expected to operate autonomously across tools and interfaces, yet production deployments require governance by construction. Systems must specify which actions are allowed, when human oversight is required, and what information may be exposed, without rebuilding the agent for each domain. This demo presents CUGA's policy system, a modular policy-as-code layer that composes with a generalist LLM agent to deliver predictable, auditable, and compliance-aware behavior in compound workflows without model fine-tuning. We present a runtime governance architecture that enforces policy interventions at every critical stage of execution. Rather than passively constraining behavior, policies intercept the agent at five structural checkpoints: upstream of planning (Intent Guard), within the system prompt to steer reasoning (Playbook), at the tool-call boundary to enforce proper usage (Tool Guide), outside the reasoning loop as a Human-in-the-Loop gate for high-risk actions (Tool Approvals), and at the output stage to filter and structure the final response (Output Formatter). Together, these stages embed governance continuously across the agent's execution pipeline rather than treating it as an afterthought. Using a healthcare scenario and a multi-layered enforcement intervention, the demo shows dynamic playbook injection for structured tool-sequence enforcement, intent guards that block malicious or accidental harmful requests, and human-in-the-loop tool approval checkpoints for potentially destructive actions. The artifact illustrates how typed governance primitives enable faster, safer deployment of enterprise agentic systems while improving policy adherence and execution consistency.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
A Reproducible Log-Driven AutoML Framework for Interpretable Pipeline Optimization in Healthcare Risk Prediction
arXiv:2605.21528v1 Announce Type: new
Abstract: Accurate and reproducible disease risk prediction remains challenging due to heterogeneous features, limited samples, and severe class imbalance. This study introduces yvsoucom-iterkit, a deterministic and log-driven automated machine learning framework that formulates pipeline optimization as a fully reproducible, configuration-level system. Each pipeline is encoded as a traceable log entity, enabling analysis of component attribution, interactions, similarity, and cross-seed robustness. Experiments on the Pima Indians Diabetes and Stroke datasets across more than 18,000 pipeline configurations reveal a structured and partially redundant search space, where performance is governed by a small subset of interacting components. Random Forest importance analysis identifies augmentation (0.454), model choice (0.198), and imbalance handling (0.101) as key drivers on Pima, while imbalance handling dominates Stroke (0.406). Component similarity analysis shows strong redundancy, with feature selection variants (biMax-biMean) exhibiting low RMS distance (0.0252), mixup closely matching no augmentation (0.0279), and TomekLinks aligning with no imbalance handling (0.0325), whereas Gaussian noise shows greater divergence from no augmentation (0.10). The framework achieves strong and stable performance using ensemble models (Weighted-F1 0.89, Macro-F1 0.88 on Pima; Weighted-F1 0.94 on Stroke), while Macro-F1 remains lower on Stroke (0.67) due to class imbalance. Cross-seed analysis reveals a performance-robustness trade-off, with ensembles showing lower variability (0.023-0.026) than SVM. These results indicate that effective AutoML optimization can focus on a reduced set of high-impact components.
Fonte: arXiv cs.LG
Vision • Score 85
Learning Emergent Modular Representations in Multi-modality Medical Vision Foundation Models
arXiv:2605.21861v1 Announce Type: new
Abstract: Multi-modality medical vision (MV) foundation models (FM) are fundamentally challenged by pronounced Non-IID feature statistics across heterogeneous imaging modalities. Monolithic self-supervised optimization on such data induces conflicting gradients, driving representations to collapse toward modality-dominant shortcuts. This work reframes this failure as an imbalance between specialization and coordination in emergent modularity, and proposes Director-Experts (DEX), a modular network that explicitly regulates these dynamics in stacked modules. Each DEX module comprises a pool of experts, dynamically adapted by our image-wise activation strategy, autonomously specializing in modality-dominant statistics, together with a director, updated via our group exponential moving average, which distills multi-expert knowledge into a shared space for semantic integration across modalities, thus driving the emergence of modular representations. We curate a new benchmark, Medical Vision Universe, over 4 million images across 10 modalities, which provides a FM-level pre-training with the broadest coverage of distinct imaging modalities to our DEX. Extensive evaluations on 26 downstream tasks demonstrate improved optimization behavior and transferability, indicating DEX as a principled step toward general-purpose multi-modality medical AI. Our code and dataset will be opened at https://github.com/YutingHe-list/DEX.
Fonte: arXiv cs.CV
Theory/Optimization • Score 85
Variance Reduction for Expectations with Diffusion Teachers
arXiv:2605.21489v1 Announce Type: cross
Abstract: Pretrained diffusion models serve as frozen teachers feeding downstream pipelines such as text-to-3D, single-step distillation, and data attribution. The teacher gradients these pipelines consume are Monte Carlo (MC) expectations over noise levels and Gaussian noise samples; their estimator variance dominates compute cost because each draw requires expensive upstream work (rendering, simulation, encoding). We introduce CARV, a compute-aware variance-accounting framework that motivates a hierarchical MC estimator: amortize the expensive upstream computation over cheap diffusion-noise resamples, sharpened by timestep importance sampling and a stratified-inverse-CDF construction. In our text-to-3D distillation and attribution experiments, CARV delivers 2-3x effective compute multipliers (most from amortized reuse; ~25% additional from IS+stratification) without changing the objective; in single-step distillation, the same techniques cut gradient variance by an order of magnitude but do not improve downstream FID, marking the regime where MC variance is no longer the bottleneck.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Federated LoRA Fine-Tuning for LLMs via Collaborative Alignment
arXiv:2605.21217v1 Announce Type: new
Abstract: Low-rank adaptation (LoRA) has emerged as a powerful tool for parameter-efficient fine-tuning of large language models (LLMs). This paper studies LoRA under a federated learning setting, enabling collaborative fine-tuning across clients while preserving parameter efficiency. We focus on a highly heterogeneous regime in which clients share only partial structure and a substantial subset may be contaminated. We propose Collaborative Low-rank Alignment and Identifiable Recovery (CLAIR), a contamination-aware framework that relies only on preliminary local estimators. Its formulation applies broadly, from linear regression to neural network and LLM modules, whenever local adaptation can be represented by matrix-valued updates. CLAIR recovers the shared LoRA subspace and detects contaminated clients via a structured low-rank plus block-sparse decomposition. We prove exact recovery of the shared LoRA subspace in the noiseless case, stable recovery under preliminary estimation error, and consistent collaborative-set recovery under mild separation conditions. We further quantify the gain from CLAIR refinement: it reduces off-subspace estimation error through cross-client averaging while preserving client-specific variation within the shared LoRA subspace, thus improves over local fine-tuning whenever this oracle gain outweighs the costs of subspace estimation and benign-client heterogeneity. Empirically, we demonstrate the benefits of CLAIR by fine-tuning a Transformer architecture on a text-copying task. The results show accurate contamination detection and improved benign-client performance compared with local fine-tuning and non-robust federated averaging.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Mix-Quant: Quantized Prefilling, Precise Decoding for Agentic LLMs
arXiv:2605.20315v1 Announce Type: new
Abstract: LLM agents have recently emerged as a powerful paradigm for solving complex tasks through planning, tool use, memory retrieval, and multi-step interaction. However, these agentic workflows often introduce substantial input-side overhead, making the compute-intensive prefilling stage a key bottleneck in long-context, multi-turn inference. In this work, we propose Mix-Quant, a simple and effective phase-aware quantization framework for fast agentic inference. We first investigate FP4 quantization in agentic LLM workflows and observe that quantizing the entire inference process can incur significant performance degradation. In contrast, the prefilling stage exhibits substantial quantization redundancy and can therefore be quantized with minimal accuracy loss, despite being the dominant source of computation. Based on this insight, we apply high-throughput NVFP4 quantization to the prefilling phase while preserving BF16 precision for decoding. By decoupling prefilling acceleration from decoding quality, Mix-Quant combines phase-aware algorithmic quantization with hardware-efficient NVFP4 execution to alleviate the inference bottleneck in LLM agents. Extensive experiments across long-context and agentic benchmarks demonstrate that Mix-Quant largely preserves task performance while delivering significant efficiency improvements, achieving up to a 3x speedup during prefilling.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
DIVE: Embedding Compression via Self-Limiting Gradient Updates
arXiv:2605.20689v1 Announce Type: new
Abstract: High-dimensional embeddings from large language models impose significant storage and computational costs on vector search systems. Recent embedding compression methods, including Matryoshka-Adaptor (EMNLP 2024), Search-Adaptor (ACL 2024), and SMEC (EMNLP 2025), enable dimensionality reduction through lightweight residual adapters, but their training objectives cause severe overfitting when labeled data is scarce, degrading retrieval performance below the frozen baseline. We propose \textsc{DIVE} (\textbf{D}imensionality reduction with \textbf{I}mplicit \textbf{V}iew \textbf{E}nsembles), a compression adapter that addresses this failure through two mechanisms. First, a self-limiting hinge-based triplet loss produces zero gradient once a triplet satisfies the margin constraint, bounding the total perturbation applied to the pretrained embedding space. Second, a head-wise NT-Xent contrastive loss treats multiple learned projections of each embedding as implicit views, providing dense self-supervised gradients that compensate for the sparsity of the triplet signal on small datasets. Across six BEIR datasets, \textsc{DIVE} outperforms all three baseline adapters on every dataset and at every evaluated compression ratio, with a 14M-parameter open-source implementation.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
GraphRAG on Consumer Hardware: Benchmarking Local LLMs for Healthcare EHR Schema Retrieval
arXiv:2605.20815v1 Announce Type: new
Abstract: Graph-based Retrieval Augmented Generation (GraphRAG) extends retrieval-augmented generation to support structured reasoning over complex corpora, but its reliability under resource-constrained, privacy-sensitive deployments remains unclear. In healthcare, where Electronic Health Record (EHR) data is complex and strictly regulated, reliance on cloud-based large language models (LLMs) introduces challenges in cost, latency, and compliance. In this work, we present a systematic evaluation of GraphRAG for EHR schema retrieval using locally deployed open-source LLMs. We implement the Microsoft GraphRAG pipeline on real-world EHR schema documentation and benchmark four models, including Llama 3.1 (8B), Mistral (7B), Qwen 2.5 (7B), and Phi-4-mini (3.8B), each deployed via Ollama on a single consumer GPU (8 GB VRAM). We evaluate indexing efficiency, knowledge graph construction, query latency, answer quality, and hallucination under both global and local retrieval modes. Our results reveal substantial differences: Llama 3.1 produces the richest knowledge graph (1,172 entities), Qwen 2.5 achieves the best answer quality (3.3/5), Phi-4-mini fails to complete the pipeline due to structured-output errors, and Mistral exhibits degenerate repetition behavior. We further show that GraphRAG exhibits a practical capacity threshold, where models below approximately 7B parameters fail to reliably produce valid structured outputs and cannot complete the pipeline. In addition, indexing and answer quality are decoupled across models, and local retrieval consistently outperforms global summarization in both latency and factual grounding, with reduced hallucination. These findings demonstrate that GraphRAG is feasible on consumer hardware while highlighting the importance of model selection and retrieval design for robust deployment in regulated settings.
Fonte: arXiv cs.CL
MLOps/Systems • Score 88
Causal Unlearning in Collaborative Optimization: Exact and Approximate Influence Reversal under Adversarial Contributions
arXiv:2605.20341v1 Announce Type: new
Abstract: Federated learning systems must support data deletion requests to comply with privacy regulations, yet retraining from scratch after each deletion is computationally prohibitive. We present HF-KCU, a method that removes a client's contribution by approximating the influence function through conjugate gradient iterations in Krylov subspaces, reducing complexity from O(d^3) to O(kd) where k<<d.A causal weighting mechanism ensures that only clients holding the deleted data receive parameter updates, preventing spurious changes to unaffected clients. Our method is designed to handle bounded adversarial perturbations to the Hessian and gradient, providing graceful degradation under realistic threat models. We validate HF-KCU across convolutional (ResNet-18, SimpleCNN) and transformer (ViT-Lite) architectures on CIFAR-10, MNIST, and Fashion-MNIST. On CIFAR-10 under Dirichlet (alpha=0.5) partitioning, HF-KCU achieves 47.75 times speedup over retraining while maintaining test accuracy within 0.60% of the rational baseline(71.16 vs 71.76 %). Membership inference attacks on the forget set yield success rates of 0.499 matching the retrained model and confirming effective privacy restoration. We provide convergence guarantees showing that the Krylov approximation error decreases as O((k ^1/2-1)/(k^1/2+1)) where k is the Hessian condition number. The causal weighting mechanism ensures surgical updates, where only clients holding deleted data are modified, preserving model quality for unaffected participants and avoiding the instability of gradient-based approaches in asynchronous federated settings. This design provides interpretability as each update is directly traceable to the influence of the deleted data. The method's efficiency and precision make it suitable for production federated systems where deletion requests arrive asynchronously and computational budgets are constrained.
Fonte: arXiv cs.LG
Privacy/Security/Fairness • Score 85
It Takes Two: Complementary Self-Distillation for Contextual Integrity in LLMs
arXiv:2605.20258v1 Announce Type: new
Abstract: Contextual Integrity (CI) defines privacy not merely as keeping information hidden, but as governing information flows according to the norms of a given context. As large language models are increasingly deployed as personal agents handling sensitive workflows, adhering to CI becomes critical. However, even frontier models remain unreliable in making disclosure decisions, and existing mitigation strategies often degrade underlying task performance. To overcome this privacy-utility trade-off, we propose SELFCI, a complementary self-distillation framework that decouples information suppression from task resolution. SELFCI jointly optimizes two independent reverse KL divergences over distinct teacher distributions derived from feedback: one encourages preserving task-relevant information for utility, while the other enforces minimal and appropriate disclosure. This complementary formulation induces a Product-of-Experts (PoE) target, aligning the policy with the intersection of capability and privacy requirements. Empirical evaluations demonstrate that SELFCI, without relying on costly external supervision, consistently outperforms competitive baselines such as online reinforcement learning algorithms (e.g., GRPO). These trends further extend to out-of-domain settings involving agentic workflows and accumulated private context, suggesting that SELFCI provides a practical path toward CI alignment.
Fonte: arXiv cs.LG
NLP/LLMs • Score 90
Quant.npu: Enabling Efficient Mobile NPU Inference for on-device LLMs via Fully Static Quantization
arXiv:2605.20295v1 Announce Type: new
Abstract: Large language models (LLMs) are increasingly deployed on mobile devices, where Neural Processing Units (NPUs) necessitate fully static quantization for optimal inference efficiency. However, existing post-training quantization (PTQ) methods predominantly rely on dynamic activation quantization, rendering them incompatible with NPU hardware constraints. To bridge the gap between high-fidelity PTQ and NPU-constrained inference, we propose Quant.npu, a integer-only fully static quantization framework. It incorporates learnable quantization parameters and rotation matrices, enabling low-bit activation-weight quantization without runtime quantization parameters re-computation. Crucially, we identify that initialization and selective optimization of quantization parameters is pivotal for optimization stability, as improper initialization and naive joint optimization induce gradient instability that disrupts the optimization of rotation matrices. To address this, we propose a rotation-and-bit-width-aware initialization tailored to diverse activation profiles and a distribution-aware selective optimization (two-stage quantization pipeline) tailored to rotated and unrotated tensors. Furthermore, we introduce a sensitivity-guided adaptive mixed-precision scheme to balance accuracy with inference efficiency. Extensive experiments on real-world mobile NPUs demonstrate that Quant.npu achieves comparable accuracy to state-of-the-art methods, while reducing inference latency by up to 15.1%.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
FusionCell: Cross-Attentive Fusion of Layout Geometry and Netlist Topology for Standard-Cell Performance Prediction
arXiv:2605.20287v1 Announce Type: new
Abstract: Standard cells form the building blocks of digital circuits, so their delay and power critically influence chip-level performance; yet characterization still relies on slow simulation sweeps, and many fast predictors ignore layout geometry, missing coupling and layout-dependent effects. The challenge is to jointly represent layout geometry and netlist topology so models capture fine-grained spatial details together with structural connectivity for accurate performance prediction. We introduce FusionCell, a dual-modality predictor that treats routed layout geometry and netlist topology as inputs and fuses them explicitly in a unified model. A DeiT encoder processes three-layer routed layouts, while a graph transformer models heterogeneous device/net graphs. The modalities are integrated through a topology-guided mechanism, where the netlist acts as a structural "map" to actively query relevant physical regions in the layout for joint geometric and topological reasoning. We build a 7nm dataset based on the ASAP7 PDK with over 19.5k cells spanning 149 types using automatic tools, targeting six metrics: signal rise/fall delay, transition, and power. Experimental results demonstrate that FusionCell reduces regression error, with an average MAPE of 0.92 percent, and improves Spearman/Kendall ranking over baselines, while accelerating the characterization process by orders of magnitude compared to circuit simulation.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
MagBridge-Battery: A Synthetic Bridge Dataset for Li-ion Magnetometry and State-of-Health Diagnostics
arXiv:2605.20240v1 Announce Type: new
Abstract: Battery health diagnostics today rely overwhelmingly on electrochemical signals measured at the cell terminals. A parallel literature has shown that magnetic sensing can resolve information that terminal-only measurements miss, but method development is limited by the absence, to the best of our knowledge, of public battery magnetic-measurement datasets paired with degradation labels. We release MagBridge-Battery v1.0, a synthetic dataset of 6,760 magnetic-field signatures that bridges real magnetic morphology from the Mohammadi-Jerschow Open Science Framework (OSF) archive with state-of-health (SOH) labels from the PulseBat dataset. The release contains 5,600 PulseBat-conditioned grounded samples, 600 synthetic sensor-anomaly samples derived from clean parents, and 560 low-voltage Regime-B extrapolation samples. A cell-disjoint, parent-child-leakage-free primary benchmark split is verified to contain zero overlapping cells, zero cross-split parent-child pairs, and zero sample-ID overlap. We define three primary benchmark tasks: SOH regression, second-life classification, and anomaly detection, plus an auxiliary anomaly-subtype classification task. A controlled label-shuffle ablation collapses SOH regression from R^2 approximately 0.77 to approximately 0, confirming that the bridge encodes input SOH non-trivially rather than producing label-aligned artifacts. The dataset is released on Zenodo under CC-BY-4.0, and the bridge code and benchmark suite are released under Apache-2.0. This work provides a public benchmark for magnetic-sensing battery diagnostics while paired magnetic-electrochemical measurements remain scarce.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
PulseCol: Periodically Refreshed Column-Sparse Attention for Accelerating Diffusion Language Models
arXiv:2605.20813v1 Announce Type: new
Abstract: Inference in diffusion large language models (dLLMs) is computationally expensive, as full self-attention must be repeatedly executed at each step of the denoising process without KV cache. Recent sparse attention methods for dLLMs mitigate this cost via block-sparse computation, which is applied only in later iterations when model performance is less sensitive to coarse-grained sparse approximation, but yields limited improvements in computational efficiency and acceleration. This motivates a finer-grained sparsification strategy that can be applied from earlier iterations and leverages reusable sparsity patterns, enabling further efficiency gains. In this work, we introduce PulseCol, a periodically refreshed column-sparse attention method for accelerating diffusion language models. PulseCol replaces coarse block-level sparsity with a finer-grained column-sparse structure, allowing important attention interactions to be retained more precisely while exposing greater sparsity. Built on this column-level formulation, PulseCol further identifies sparse patterns at the early denoising step and reuses them across subsequent iterations, refreshing them only at a small number of intermediate steps to track the evolution of sparse attention patterns during denoising. Experiments show that PulseCol achieves higher sparsity and greater practical speedup than prior sparse attention methods for dLLMs, while maintaining model quality. Enabled by optimized GPU kernels for column-sparse attention, PulseCol delivers up to 1.95$\times$ end-to-end speedup over FlashAttention across several context lengths.
Fonte: arXiv cs.CL
NLP/LLMs • Score 90
Introspective X Training: Feedback Conditioning Improves Scaling Across all LLM Training Stages
arXiv:2605.20285v1 Announce Type: new
Abstract: We tackle the question of how to scale more efficiently across the many, ever-growing stages of current LLM training pipelines. Our guiding intuition stems from the fact that the dynamics of later stages of the pipeline, e.g. post-training, can be used to inform earlier stages such as pre-training. To this end, we propose Introspective Training (or IXT), inspired by offline reward-conditioned reinforcement learning and applicable to any stage of training. IXT uses a thinking reward model to annotate data with natural language critique based feedback, enabling quality aware training from the earliest stages of the pipeline. Models are then trained by prefix-conditioning the data with the generated feedback -- ensuring that not all tokens are treated equally starting much earlier in training than usual. Comprehensive experiments on 7.5-12B transformer-based dense LLMs trained from scratch all the way up to 18 Trillion tokens seen show that our method: bends scaling curves resulting in up to 2.8x more compute efficiency generally; and reaches performance levels unachievable for models trained otherwise in domains such as math and code.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Modality-Decoupled Online Recursive Editing
arXiv:2605.20273v1 Announce Type: new
Abstract: Online model editing for multimodal large language models (MLLMs) requires assimilating a stream of corrections under tight compute and memory budgets. Yet editors developed for text-only LLMs often degrade on MLLMs: visually dominant activations skew the statistics that shape updates, causing cross-modal conflict, while sequential writes become entangled in a shared edit space and amplify long-horizon interference, causing inter-edit interference. To address these, we propose M-ORE, a modality-decoupled online recursive editor for lifelong MLLM adaptation. M-ORE is derived from a unified proximal-projection formulation and admits a closed-form update with a Sherman-Morrison recursion, yielding constant per-edit overhead. It maintains module-wise locality statistics for the text stack and the visual projector to avoid visually dominated update shaping and performs continual updates in a fixed orthogonal low-rank edit subspace via a Sherman-Morrison recursion to mitigate long-horizon interference. Experiments on multiple MLLM backbones and online editing benchmarks show that our M-ORE method consistently improves reliability, generality, and locality over strong baselines, while achieving favorable quality-efficiency scaling. Our code is publicly available at https://github.com/lab-klc/M-ORE.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Conformal Selective Acting: Anytime-Valid Risk Control for RLVR-Trained LLMs
arXiv:2605.20270v1 Announce Type: new
Abstract: A local specialist LLM, fine-tuned with reinforcement learning from verifiable rewards (RLVR) on operator-local data, is installed in a regulated organization with per-deployment error budget $\alpha$. The operator needs a safety certificate for this deployment's stream at every round: no pooling across deployments, no waiting for a long-run average. Existing wrappers cannot deliver this on adaptive, online-updated streams: offline conformal-risk methods require exchangeability; online-conformal methods bound only long-run averages; non-exchangeable extensions are marginally valid; and the closest anytime wrapper, A-RCPS, controls marginal rather than selective risk. Using a (test statistic, validity guarantee, deployment rule) framework, we identify one empty cell forced by deployment requirements: e-process per threshold, selective risk, anytime-pathwise validity, max-certified-threshold rule. Conformal Selective Acting (CSA) fills it as a per-round wrapper maintaining a Ville-type e-process per threshold on a Bonferroni grid, evaluated against the RLVR filtration. Under predictable updates and isotonic-calibrated monotone risk we prove (i) an anytime-pathwise selective-risk bound $R_T^{\mathrm{act}}\le\alpha+O(N_T^{-1/2})$, (ii) rate-optimal certification matching $\Theta(\bar\eta^{-2}\log(1/\delta))$, and (iii) a horizon-independent release-rate gap. Across eight specialist benchmarks ($480$ streams), sixteen adversarial distribution-shift cells ($160$ streams), and five live Expert-Iteration RLVR cells with online LoRA over four base models in three architecture families ($10{,}300$ rounds), CSA is the only method among ten compared that satisfies pathwise validity and non-refusing deployment on every cell. We do not propose a new LLM, training algorithm, or policy class; CSA is the deployment-side complement, orthogonal to the model, for operators who cannot use a frontier API.
Fonte: arXiv cs.LG
RecSys • Score 85
Group-Aware Matrix Estimation and Latent Subspace Recovery
arXiv:2605.20559v1 Announce Type: new
Abstract: Modern matrix completion problems often involve heterogeneous data whose rows simultaneously belong to many meta-categories, such as demographic and age groups in recommendation systems, or region and recording session labels in neural electrophysiological experiments. Standard low-rank estimators impose a single global latent geometry, which can recover average structure but may smooth away subgroup-specific variation, especially when observations are unevenly distributed across groups. We introduce Group-Aware Matrix Estimation (GAME), a convex estimator for overlapping subgroup-wise low-rank matrix estimation. GAME regularizes category-specific submatrices through overlapping nuclear-norm penalties, allowing related groups to borrow information while preserving local latent structure in a shared coordinate system. We provide finite-sample guarantees for both reconstruction error and subgroup-specific subspace recovery, showing how performance depends on sampling density, subgroup rank, and overlap structure. Experiments on synthetic, recommendation, ecological, and neuroscience datasets show that GAME is most beneficial in structured missingness regimes, where subgroup-aware regularization improves both reconstruction accuracy and latent subspace fidelity. Across these benchmarks, GAME is competitive or best among global low-rank, side-information, and modern imputation baselines, with the largest gains when subgroups exhibit distinct low-rank structure.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Plug-and-Play Spiking Operators: Breaking the Nonlinearity Bottleneck in Spiking Transformers
arXiv:2605.20289v1 Announce Type: new
Abstract: ANN-to-SNN conversion offers a practical, training-free route to spiking large language models. However, current pipelines primarily focus on spike-driven realizations for Transformer linear-algebra operations, while providing limited support for key nonlinear operators. This gap limits compatibility with neuromorphic-style execution constraints, where such nonlinearities typically require division, exponentiation, or norm computations that are not naturally supported by standard leaky integrate-and-fire dynamics. To solve this problem, we propose a plug-and-play framework that implements spike-friendly approximations for Transformer nonlinearities and integrates into existing ANN-to-SNN pipelines. Our method decomposes these nonlinear computations into three recurring primitives -- division, exponentiation, and $\ell_2$ norms -- and realizes them via population computation using LIF neuron groups, combined with lightweight bit-shift scaling to avoid floating-point arithmetic. By composing these primitives as modular operator blocks, our framework supports common Transformer nonlinearities (e.g., Softmax, SiLU, and normalization) without any fine-tuning. Experiments on a range of LLMs Transformers show that selectively replacing the targeted nonlinear operators incurs less than a $1\%$ accuracy drop across all evaluated tasks.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Less Data, Faster Training: repeating smaller datasets speeds up learning via sampling biases
arXiv:2605.20314v1 Announce Type: new
Abstract: This work investigates the ``small-vs-large gap'', where repeating on fewer samples can lead to compute saving during training compared to using a larger dataset. This is observed across algorithmic tasks, architectures and optimizers and cannot be explained using prior theory. We argue that the speedup comes from appropriate layer-wise growth enabled by sampling biases, which is more pronounced when the dataset size is smaller. We provide both theoretical analysis and empirical evidence from various interventions. Our results suggest that using a smaller dataset with more repetitions is not just a fallback strategy under data scarcity, but can be proactively leveraged as a favorable inductive biases for optimization, particularly in reasoning tasks.
Fonte: arXiv cs.LG
RL • Score 85
GROW: Aligning GRPO with State-Action Modeling for Open-World VLM Agents
arXiv:2605.20246v1 Announce Type: new
Abstract: Recently, vision-language model (VLM) agents have shown promising progress in open-world tasks, where successful task completion often requires multiple turns of visual perception and action execution. However, existing methods still rely primarily on Supervised Fine-Tuning (SFT) with expert demonstrations, while the advanced reinforcement learning (RL) algorithm, specifically Group Relative Policy Optimization (GRPO), has not been effectively employed for multi-turn RL in these tasks because standard GRPO requires full trajectories as training samples which leads to excessively long context and noise. To address this issue, we propose GROW, a RL framework for open-world VLM agents that decomposes collected trajectories into state-action samples, and computes advantages between these samples rather than treating a full trajectory as a single entity. We further provide a surrogate analysis indicating that, even though the grouped samples are conditioned on different local states rather than an identical prompt context, the objective can preserve the core relative policy optimization signal of GRPO under simplifying assumptions. Experiments on more than 800 Minecraft tasks show that our method achieves state-of-the-art (SOTA) performance, demonstrating the effectiveness of our proposed RL framework for open-world VLM agents.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Weasel: Out-of-Domain Generalization for Web Agents via Importance-Diversity Data Selection
arXiv:2605.20291v1 Announce Type: new
Abstract: Large language models (LLMs) have enabled web agents that follow natural language goals through multi-step browser interactions. However, agents fine-tuned on specific trajectories and domain often struggle to generalize out of domain, and offline training can be compute-inefficient due to noisy, redundant trajectories and long accessibility-tree (AXTree) states. To address both issues, we propose Weasel, a trajectory selection method for offline training of web agents. Weasel selects a fixed-budget subset of trajectory steps by optimizing an objective that balances unary importance with pairwise diversity over states, websites, and interaction patterns, solving efficiently with a greedy algorithm. We further improve efficiency with target-centered AXTree pruning that keeps only content around the ground-truth action target, and we mitigate style mismatch for reasoning-native models by replacing expert traces with model-generated, style-consistent rationales. Across AgentTrek and NNetNav training datasets, evaluations in WebArena, WorkArena, and MiniWob, and experiments with Qwen2.5-7B, Gemma3-4B, and Qwen3-8B, Weasel improves out-of-domain performance while reducing training cost, producing roughly 9.7-12.5$\times$ training speedups over standard fine-tuning. We make the code available at https://github.com/fatemehpesaran310/weasel.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
OmniISR: A Unified Framework for Centralized and Federated Learning via Intermediate Supervision and Regularization
arXiv:2605.20276v1 Announce Type: new
Abstract: The global deployment of edge intelligence operates across heterogeneous legal frameworks. While some regions permit centralized learning (CL) via cloud data aggregation, others enforce strict data localization, necessitating federated learning (FL). This operational dichotomy introduces two incompatible optimization regimes (i.e., unbiased global gradients yet coupled with internal covariate shift in CL versus biased, drift-prone local updates in FL), resulting in that any naive integration of the two lacks rigorous theoretical guarantees. To fill this gap, we propose OmniISR, a unified framework that fuses pure CL, pure FL, and hybrid CL-FL training modes via equipping intermediate supervision and regularization (ISR) signals at multiple hidden layers. Specifically, we propose (i) to use mutual-information (MI) as intermediate supervision to align shifting internal covariate in CL and client-drifting representations in FL, and (ii) to adopt negative-entropy (NE) as intermediate regularizer to penalize overconfident prediction, preserve representational uncertainty, and avoid device-specific collapse. On the theory side, we derive (i) a unified, ISR-agnostic, and non-asymptotic O(1/sqrt(T)) convergence bound that shows the introduced ISR does not violate standard SGD convergence, (ii) a federated drift-bound that quantifies the ISR-reduced client drift, (iii) a gradient-alignment guarantee that ensures non-conflicting CL and FL updates under mild bias, and (iv) an explicit escape-time bound that indicates that CL-FL hybrid mixing enlarges effective stochasticity and accelerates escape from strict saddles. Extensive experiments demonstrate that OmniISR consistently improves model performance in both centralized and federated paradigms, reduces the CL-FL gap by 22.60%, and yields 37/48 paired metric wins across multiple FL algorithms.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Auto-Dreamer: Learning Offline Memory Consolidation for Language Agents
arXiv:2605.20616v1 Announce Type: new
Abstract: Language agents increasingly operate over streams of related tasks, yet existing memory systems struggle to convert accumulated experience into reusable knowledge. Retrieval-augmented and structured memory methods record per-session observations effectively, but often couple acquisition and consolidation into a single online process, leaving the agent without a global view across sessions to discover recurring patterns, abstract shared procedures, or prune redundant entries. Inspired by complementary learning systems theory, we propose Auto-Dreamer, a learned offline consolidator for language-agent memory. Auto-Dreamer decouples fast per-session memory acquisition from slow cross-session consolidation. Given a selected working region of a typed memory bank, the consolidator treats the region as read-only evidence, performs bounded tool-use to inspect entries and provenance-linked source trajectories, and synthesizes a fresh compact replacement set that abstracts across sessions and supersedes the original region. We train Auto-Dreamer via GRPO, using end-to-end agent performance as the reward signal to learn how to consolidate memories acquired through fast online experience. Trained on ScienceWorld trajectories alone, Auto-Dreamer outperforms fixed, RL-trained, and prompted memory baselines on ScienceWorld by 7 points while using an active memory bank 12$\times$ smaller than the strongest baseline, and continues to lead on held-out ALFWorld and WebArena without retraining -- using 6$\times$ less memory than the strongest baseline on ALFWorld.
Fonte: arXiv cs.CL
MLOps/Systems • Score 85
LOSCAR-SGD: Local SGD with Communication-Computation Overlap and Delay-Corrected Sparse Model Averaging
arXiv:2605.20866v1 Announce Type: cross
Abstract: Communication is a major bottleneck in distributed learning, especially in large-scale settings and in federated learning environments with slow links. Three standard ways to reduce this cost are communication compression, local training, and communication-computation overlap. Methods that combine these ingredients are used in practice and have been found to be effective for large-scale training, but there is little theory for methods that combine all three. We study a heterogeneous-compute setting in which different workers may take different numbers of local steps, and we propose LOSCAR-SGD, a Local SGD method that communicates only a sparse subset of model coordinates and continues optimizing while communication is in flight. A key ingredient is a delay-corrected merge rule that incorporates delayed synchronized information without discarding the progress made during the overlap phase. We give convergence guarantees for smooth non-convex objectives and show how sparsity, overlap, and worker heterogeneity affect the rate. To the best of our knowledge, this is the first theory for this combination of ingredients. Experiments further show that communication-computation overlap reduces training time and that the delay-corrected merge outperforms naive overwriting.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Terminal-World: Scaling Terminal-Agent Environments via Agent Skills
arXiv:2605.20876v1 Announce Type: new
Abstract: Terminal agents extend Large Language Models with the ability to execute tasks directly in command-line environments, but their progress is bottlenecked by the scarcity of high-quality training data. Existing approaches bootstrap from partial sources such as human-defined seeds or GitHub repositories to instantiate one component and then complete the rest, producing tasks confined to narrow seed distributions, environments misaligned with task semantics, and inefficient trajectories from unguided exploration. To address these limitations, we introduce Terminal-World, a fully automated pipeline that uses agent skills as the central synthesis primitive, which jointly encode what to accomplish, when to apply (preconditions and environment state), and how to execute, enabling task instructions, environments, and teacher trajectories to be co-derived. To further broaden the synthesis space, Terminal-World composes skills into skill teams and skill graphs for multi-role and cross-domain task synthesis. Using this pipeline, we construct 5,723 training environments and train Terminal-World-8B/14B/32B, evaluated across 6 benchmarks where the Terminal-World series consistently outperforms terminal-agent baselines. Notably, using the same teacher model and only 1.2% of the training data, Terminal-World-32B surpasses Nemotron-Terminal-32B on Terminal-Bench 2.0 by +4.5 Pass@1 (31.5) and achieves 43.8 Pass@3.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
Learning to Hand Off: Provably Convergent Workflow Learning under Interface Constraints
arXiv:2605.19140v1 Announce Type: new
Abstract: We study workflow learning in a setting where specialized agents hand off control through a shared artifact, each agent observes only a local function of that artifact and its own private state, and no centralized learner accesses joint trajectories -- the operating regime of multi-agent LLM pipelines that span organizational, vendor, or trust boundaries. We formalize this regime as an interface-constrained semi-Markov decision process (IC-SMDP), whose decision epochs occur at handoff times, and design IC-$Q$, an asynchronous decentralized $Q$-learning algorithm in which cross-agent coordination at every handoff is exactly one scalar. Our main result is a finite-sample bound for neural IC-$Q$ that decomposes into three independently controllable error sources: neural function-approximation error, interface representation gap, and a mixing-time residual, under the random option-duration discount. Establishing this bound requires lifting the approximate information state (AIS) framework from single-agent primitive-step MDPs to multi-agent SMDPs and controlling Markovian noise under random duration, neither of which has been done in prior work. To our knowledge this is the first finite-sample guarantee for neural $Q$-learning under decentralized partial observability. Four experiments: a controlled synthetic IC-SMDP that validates the bound term-by-term, multi-LLM mathematical reasoning, multi-agent routing, and multi-agent CPU programming, show that IC-$Q$ matches a centralized oracle without any agent observing joint trajectories, with each of the three error sources scaling along its corresponding axis as the bound predicts.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
MiMuon: Mixed Muon Optimizer with Improved Generalization for Large Models
arXiv:2605.19619v1 Announce Type: cross
Abstract: Matrix-structured parameters frequently appear in many artificial intelligence models such as large language models. More recently, an efficient Muon optimizer is designed for matrix parameters of large-scale models, and shows markedly faster convergence than the vector-wise algorithms. Although some works have begun to study convergence properties (i.e., optimization error) of the Muon optimizer, its generalization properties (i.e., generalization error) is still not established. Thus, in this paper, we study generalization error of the Muon optimizer based on algorithmic stability and mathematical induction, and prove that the Muon has a generalization error of $O\big(\frac{1}{N\kappa^{T}}\big)$, where $N$ is training sample size, and $T$ denotes iteration number, and $\kappa>0$ denotes minimum difference between singular values of gradient estimate. To enhance generalization of the Muon, we propose an effective mixed Muon (MiMuon) optimizer by cautiously using orthogonalization of gradient, which is a hybrid of Muon and momentum-based SGD optimizers. Then we prove that our MiMuon optimizer has a lower generalization error of $O\big(\frac{1}{N}\big)$ than $O\big(\frac{1}{N\kappa^{T}}\big)$ of Muon optimizer, since $\kappa$ generally is very small. Meanwhile, we also studied the convergence properties of our MiMuon algorithm, and prove that our MiMuon algorithm has the same convergence rate of $O(\frac{1}{T^{1/4}})$ as the Muon algorithm. Some numerical experimental results on training large models including Qwen3-0.6B and YOLO26m demonstrate efficiency of the MiMuon optimizer.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Exact Linear Attention
arXiv:2605.18848v1 Announce Type: new
Abstract: This paper introduces Exact Linear Attention (ELA), a mechanism that achieves linear computational complexity for Transformer attention by leveraging the exact decomposition property of kernel functions, without any approximation error. It identifies and addresses gradient explosion and token attention dilution in prior linear attention methods by imposing kernel constraints that ensure non-negativity, discriminability, and geometric interpretability. Several kernel functions are proposed, including the Hadamard Exp Kernel, Summation Squared Euclidean Distance Kernel, and Subtraction Squared Euclidean Distance Kernel. Beyond the core attention formulation, the paper presents three engineering innovations: a Hyper Link structure that replaces traditional residual connections to mitigate gradient degradation, a Memory Lobe module based on bidirectional linear attention that captures transformation flow across layers to implement qualitative memory and an implicit reinforcement learning paradigm, and a routing score based bias mechanism for Mixture of Experts to improve interpretability and semantic alignment.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Interference-Aware Multi-Task Unlearning
arXiv:2605.19042v1 Announce Type: new
Abstract: Machine unlearning aims to remove the contribution of designated training data from a trained model while preserving performance on the remaining data. Existing work mainly focuses on single-task settings, whereas modern models often operate in multi-task setups with shared backbones, where removing supervision for one task or instance can unintentionally affect others. We introduce multi-task unlearning with two settings: full-task unlearning, which removes a target instance from all tasks, and partial-task unlearning, which removes supervision only from selected tasks. We show that shared parameters couple the forget and retain sets, causing task-level interference on non-target tasks and instance-level interference on other instances. To address this issue, we propose an interference-aware framework that combines task-aware gradient projection, which constrains updates within task-specific subspaces, with instance-level gradient orthogonalization, which reduces conflicts between forget and retain signals. Experiments on two multi-task computer vision benchmarks across five tasks show that our method achieves effective unlearning while maintaining strong generalization, reducing UIS compared with the strongest baseline by 30.3% in full-task unlearning and 52.9% in partial-task unlearning.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Gaussian Approximation and Multiplier Bootstrap for Federated Linear Stochastic Approximation
arXiv:2605.19629v1 Announce Type: new
Abstract: In this paper, we establish Berry-Esseen-type bounds for federated linear stochastic approximation (LSA). Our results provide the first federated Gaussian approximations for LSA that explicitly capture communication-computation trade-offs and heterogeneity-aware error terms, quantifying the effects of local step size, number of local updates, and heterogeneity on convergence rates. We present results for both (i) constant step size regime and (ii) decreasing step size with an increasing number of local iterations, recovering the recent rates of Bonnerjee et al. [2025] as a special case. As a primary application of our results, we develop an online multiplier bootstrap procedure for inference on the last iterate, which avoids explicit estimation of the asymptotic covariance matrix, and obtain non-asymptotic validity guarantees for this procedure.
Fonte: arXiv stat.ML
Evaluation/Benchmarks • Score 90
OpenCompass: A Universal Evaluation Platform for Large Language Models
arXiv:2605.19276v1 Announce Type: new
Abstract: In recent years, the field of artificial intelligence has undergone a paradigm shift from task-specific small-scale models to general-purpose large language models (LLMs). With the rapid iteration of LLMs, objective, quantitative, and comprehensive evaluation of their capabilities has become a critical link in advancing technological development. Currently, the mainstream static benchmark dataset-based evaluation methods face challenges such as the diversity of task types, inconsistent evaluation criteria, and fragmentation of data and processing workflows, making it difficult to efficiently conduct cross-domain and large-scale model evaluation. To address the aforementioned issues, this paper proposes and open-sources OpenCompass, a one-stop, scalable, and high-concurrency-supported general-purpose LLM evaluation platform. Adhering to the design philosophy of modularization and component decoupling, the platform boasts three core advantages: high compatibility, flexibility, and high concurrency. The core architecture of OpenCompass comprises five key components: the Configuration System, Task Partitioning Module, Execution and Scheduling Module, Task Execution Unit, and Result Visualization Module. Its workflow provides rule-based, LLM-as-a-Judge, and cascaded evaluators to adapt to the requirements of different task scenarios. Supporting mainstream benchmark datasets across multiple domains, including knowledge, reasoning, computation, science, language, code, etc., the platform offers a unified and efficient LLM evaluation tool for both academia and industry, facilitating the accurate identification of strengths and weaknesses of LLMs as well as their subsequent optimization.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
ScheduleFree+: Scaling Learning-Rate-Free & Schedule-Free Learning to Large Language Models
arXiv:2605.19095v1 Announce Type: cross
Abstract: Schedule-Free Learning has shown promise as a practical anytime training method for machine learning, showing success across dozens of standard benchmark problems. However, strong performance for LLM training has only been demonstrated at small scales. We identify a number of fixes necessary to scale up Schedule-Free Learning to larger batch sizes and model sizes, and present a learning-rate-free and schedule-free method (ScheduleFree+) for training large language models which greatly outperforms Warmup-Stable-Decay (WSD) schedules. We also demonstrate that Schedule-Free Learning is most effective for long duration training, and at 1000 tokens per parameter, it outperforms SOTA schedules by 31%. Schedule-Free Learning provides a theoretical foundation for the use of model averaging and checkpoint merging during pretraining.
Fonte: arXiv stat.ML
MLOps/Systems • Score 85
Learn-by-Wire Training Control Governance: Bounded Autonomous Training Under Stress for Stability and Efficiency
arXiv:2605.19008v1 Announce Type: new
Abstract: Modern language-model training is increasingly exposed to instability, degraded runs, and wasted compute, especially under aggressive learning-rate, scale, and runtime-stress conditions. This paper introduces Learn-by-Wire Guard (LBW-Guard), a bounded autonomous training-control governance layer that operates above AdamW. Rather than replacing the optimizer update rule, LBW-Guard observes training telemetry, interprets instability-sensitive regimes, and applies bounded control to optimizer execution while preserving fixed training objectives.
We evaluate LBW-Guard in a Qwen2.5-centered stress-and-robustness suite using WikiText-103, with Qwen2.5-7B as the empirical anchor, model-size comparisons against Qwen2.5-3B and Qwen2.5-14B, learning-rate stress tests, gradient-clipping baselines, and a no-LoRA TinyLlama-1B full-parameter sanity check. In the 7B reference setting, LBW-Guard reduces final perplexity from 13.21 to 10.74, an 18.7% improvement, while reducing end-to-end time from 392.54s to 357.02s, a 1.10x speedup. Under stronger learning-rate stress, AdamW degrades to 1885.24 final perplexity at LR=3e-3 and 659.76 at LR=1e-3, whereas LBW-Guard remains trainable at 11.57 and 10.33, respectively. Gradient-clipping baselines do not reproduce this effect.
These results support a scoped systems conclusion that stability-sensitive LLM training can benefit from a governance plane above the optimizer. LBW-Guard provides evidence that bounded runtime control can preserve productive compute under stress while remaining distinct from optimizer replacement and local gradient suppression.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Block-Based Double Decoders
arXiv:2605.18807v1 Announce Type: new
Abstract: Encoder-decoder models offer substantial inference-time savings over decoder-only models, but their pretraining objectives suffer from sparse supervision and dynamic sequence lengths, keeping them out of practice at scale. We propose block-based double decoders, a novel transformer architecture that utilizes doubly-causal block-based attention masks to train with full loss supervision and static sequence packing, combining decoder-only training efficiency with encoder-decoder inference efficiency. In scaling law experiments, block-based double decoders strongly outperform encoder-decoders and closely track decoder-only models across scales. At inference time, they cut KV-cache memory and per-token compute by at least 2/3 without sacrificing prefill caching or other existing inference optimizations available to decoder-only models.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
ContextRAG: Extraction-Free Hierarchical Graph Construction for Retrieval-Augmented Generation
arXiv:2605.19735v1 Announce Type: new
Abstract: Graph-structured retrieval-augmented generation (RAG) systems can improve answer quality on multi-hop questions, but many current systems rely on large language models (LLMs) to extract entities, relations, and summaries during indexing. These calls add token and wall-clock costs that grow with corpus size. We present ContextRAG, a graph RAG system whose graph topology is constructed without LLM-based entity or relation extraction. ContextRAG derives a fuzzy concept graph over chunk embeddings using residual-quantization k-means and Formal Concept Analysis with Lukasiewicz residuated logic. Bridge-like and meet-derived context nodes are induced by soft fuzzy join and meet operations, rather than by LLM-written graph edges. On a 130-task UltraDomain subset, ContextRAG builds its index with 30 LLM calls and 22,073 tokens. In contrast, a local HiRAG reproduction stress test required 870 indexing calls and 3.54M tokens on a 20-task subset before failing during graph construction; linear extrapolation to 130 tasks implies over 23M indexing tokens. ContextRAG obtains 33.6% F1 overall and 36.8% F1 on multi-hop tasks. An activation analysis shows that queries retrieving at least one lattice-derived node in the top five achieve +3.9 percentage points F1 over queries that do not; this association is diagnostic rather than causal.
Fonte: arXiv cs.CL
Theory/Optimization • Score 85
Information Processing Capacity of Stationary Physical Systems: Theory, Data-efficient Estimation Methods, and Photonic Demonstration
arXiv:2605.19152v1 Announce Type: new
Abstract: Physical computing systems provide a promising route toward hardware-native machine learning, but their computational capabilities remain difficult to characterize in a principled, task-independent, and data-efficient way. We extend the Information Processing Capacity (IPC) framework to stationary physical computing systems and establish several fundamental results: individual capacities are bounded between zero and one, their sum over a complete basis is bounded by the number of readouts, and noise strictly reduces this bound. We address the finite-sample estimation of IPC and derive the asymptotic form of the systematic positive bias affecting naive estimators. Building on these results, we introduce data-efficient estimation methods based on Richardson extrapolation and Sobol quasi-random sampling. We validate the framework experimentally using a photonic computing system based on picosecond laser pulses propagating through a nonlinear optical fibre. By varying the laser power and fibre length, we observe systematic shifts of the IPC distribution toward higher-order nonlinear capacities induced by the Kerr effect. Finally, we demonstrate that the total IPC strongly correlates with performance on benchmark machine-learning tasks and provides a reliable estimate of the effective dimensionality of the system. These results establish IPC as a practical bridge between the intrinsic dynamics of physical computing systems and their machine-learning performance.
Fonte: arXiv stat.ML
MLOps/Systems • Score 85
Multi-Pedestrian Safety Warning at Urban Intersections Use Case of Digital Twin
arXiv:2605.18823v1 Announce Type: new
Abstract: Digital twins (DTs) for urban transportation systems have gained increasing attention; however, their systematic evaluation in safety-critical scenarios remains limited. This paper presents a multi-pedestrian safety warning system at urban intersections enabled by a tightly coupled physical-digital twin framework. Built upon the COSMOS city-scale wireless testbed in New York City, the proposed system integrates camera and ultra-wideband (UWB), edge-cloud computing, predictive trajectory modeling, and MQTT-based communication to deliver real-time safety alerts to vulnerable road users (VRUs). The system is evaluated through both field deployment and virtual reality (VR) experiments. Results demonstrate high warning generation accuracy, localization accuracy, efficient end-to-end latency under different model configurations, and significant reductions in user response time when warnings are issued. The proposed DT framework provides a scalable, modular, and generalizable solution for real-time multi-pedestrian safety enhancement at complex urban intersections.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
TextAlign: Preference Alignment for Text Rendering with Hierarchical Rewards
arXiv:2605.19320v1 Announce Type: new
Abstract: Faithful text rendering remains a persistent weakness of large text-to-image generative models, as it requires both semantic instruction following and fine-grained glyph-level structure. Prior methods often improve this ability through architecture-specific modules or encoder modifications, which complicate deployment across foundation models. We study text rendering as a post-training preference-alignment problem and propose TextAlign, a non-invasive framework that keeps the generator architecture unchanged. The key component is a hierarchical vision-language model (VLM)-based reward that decomposes rendering errors into global, word, and glyph levels, then converts binary defect judgments into a scalar preference signal. The resulting signal supports both Group Relative Policy Optimization (GRPO) and Direct Preference Optimization (DPO). Experiments on FLUX.1-dev and Z-Image-Turbo show consistent gains in OCR-based text accuracy without degrading general generation quality. Compared with strong foundation and text-rendering baselines, including SD3.5, Qwen-Image, AnyText, and TextDiffuser, these results indicate that reward design offers a scalable alternative to model redesign for improving text rendering.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Quantized Machine Learning Models for Medical Imaging in Low-Resource Healthcare Settings
arXiv:2605.19207v1 Announce Type: new
Abstract: Deep learning models have shown strong performance in medical image analysis, but deploying them in low-resource clinical environments remains difficult due to computational, memory, and power constraints. This paper presents a multi-strategy compression framework for brain tumor classification from MRI, encompassing quantization-aware training, knowledge distillation from a DenseNet-101 teacher to a compact DenseNet-32 student with low-bit post-training quantization, and Float16 post-training quantization on a lightweight MobileNetV2 backbone.
Using a multi-class brain tumor MRI dataset containing glioma, meningioma, pituitary tumors, and healthy controls, we provide full experimental validation of the MobileNetV2-based pipeline, training the classifier through a three-stage transfer learning process and applying Float16 quantization via TensorFlow Lite. The DenseNet-based distillation and quantization-aware training strategies are described as complementary compression approaches within the framework, with their complete empirical evaluation reserved for future work.
Experimental results on the MobileNetV2 pipeline show that the quantized model achieves 82.37 percent validation accuracy compared to the 82.20 percent full-precision baseline, reducing model size from 35.34 MB to 5.76 MB, a 6.14x compression ratio with no meaningful accuracy loss. Per-class evaluation confirms that quantization preserves diagnostic performance uniformly across all four tumor categories. These findings demonstrate that lightweight quantized models can deliver clinically viable brain tumor screening in resource-constrained healthcare settings.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
m3BERT: A Modern, Multi-lingual, Matryoshka Bidirectional Encoder
arXiv:2605.19568v1 Announce Type: new
Abstract: Embedding models are pivotal in industrial information retrieval systems like search and advertising. However, existing pretrained models often exhibit fixed architectures and embedding dimensionalities, posing significant challenges when adapting them to diverse deployment scenarios with varying business-driven constraints. A common practice involves fine-tuning with partial parameter initialization from larger pretrained models for resource-constrained tasks. This method is often suboptimal as the misalignment between pretraining and downstream usage prevents full realization of pretraining benefits. To address this limitation, we introduce m3BERT: a Modern, Multi-lingual, Matryoshka Bidirectional Encoder, which features a novel pretraining strategy that jointly optimizes representations across both transformer layers and multiple embedding dimensions. This enables a single model to be tailored to varied resource and accuracy targets while maintaining consistency with pretraining. Incorporating recent architectural improvements, m3BERT uses a three-stage pretraining: monolingual pretraining, multilingual adaptation to serve diverse user bases, and crucial continual pretraining on a massive web domain corpus to enhance utility in commercial retrieval. m3BERT significantly outperforms state-of-the-art embedding models in Bing-Click, a large-scale industrial retrieval dataset, showcasing its practical versatility as an efficient foundation for resource-aware industrial retrieval systems. Further experiments on public datasets also confirm the general effectiveness of our multigranular Matryoshka pretraining strategy.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
Hybrid-LoRA: Bridging Full Fine-Tuning and Low-Rank Adaptation for Post-Training
arXiv:2605.18822v1 Announce Type: new
Abstract: Post-training has become essential for adapting large language models (LLMs) to complex downstream behaviors, including instruction following, preference alignment, and multi-step reasoning. Reinforcement learning with verifiable rewards (RLVR) has recently emerged as a particularly effective post-training paradigm for improving reasoning capabilities, with critic-free algorithms such as GRPO and GSPO enabling scalable optimization. However, RLVR post-training with full fine-tuning (FFT) requires substantial GPU memory and incurs high training costs. Although parameter-efficient fine-tuning (PEFT) methods, such as Low-Rank Adaptation (LoRA), effectively reduce computational costs, they often suffer from a noticeable performance gap compared to full fine-tuning in post-training for complex reasoning tasks. In this paper, we propose Hybrid-LoRA, an efficient hybrid post-training framework that selectively applies full fine-tuning to a small subset of modules less suited to low-rank adaptation, while adapting the remaining components with LoRA. We introduce a novel Hybrid-LoRA Score to rank candidate modules according to their sensitivity to low-rank adaptation under a fixed parameter budget. Experiments show that Hybrid-LoRA closely matches full fine-tuning performance under a 10% full fine-tuning module budget, with the remaining candidate modules adapted by LoRA, consistently outperforming four state-of-the-art PEFT post-training baselines, achieving improvements of up to 5.65% and on average 4.36% over the best baseline.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Simply Stabilizing the Loop via Fully Looped Transformer
arXiv:2605.18797v1 Announce Type: new
Abstract: Scaling model performance typically requires increasing model size. Looped Transformer offers a compelling alternative by iteratively reusing the same Transformer blocks, trading additional computation for improved performance without increasing parameter count or context length. Because the number of loop iterations can be adjusted at inference, it also provides a natural mechanism for balancing performance and test-time compute. However, Looped Transformer still suffers from training instability when the number of loop iterations increases. Our analysis reveals that this instability stems from two sources: gradient oscillation and residual explosion. To address these two problems, we propose the Fully Looped Transformer, which introduces two parameter-free modifications: (1) Fully Looped Architecture, which distributes inter-loop signals across all layers to mitigate residual explosion; (2) Attention Injection, which reuses the existing attention block to suppress gradient oscillation. These modifications stabilize training dynamics, enabling the Fully Looped Transformer to be trained stably up to 12 loop iterations, whereas other baseline looped models collapse in this regime. In milder settings where Looped Transformer does not collapse, Fully Looped Transformer still improves average downstream-task performance by up to 13.2\%. Overall, our experiments demonstrate that Fully Looped Transformer improves training stability, enhances downstream performance, and provides preliminary adaptability under different test-time compute budgets by varying loop iterations at inference.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Graph-Driven Cross-Industry Real-Time Monitoring Framework for Anti-Money Laundering Detection in Converged Mobility-Energy Supply Chain Networks
arXiv:2605.18844v1 Announce Type: new
Abstract: With the deep integration of the travel and energy industries, cross-industry supply chain finance has gradually become a high-risk field of hidden money laundering incidents. For this reason, this work proposes a graph-driven cross-industry real-time anti-money laundering monitoring framework (GCRMF) for integrated travel - energy supply chain networks. First, a cross-industry heterogeneous graph (CIHG) covering new energy vehicle rental platforms, energy suppliers, fintech institutions, etc., is constructed, and industry semantics are integrated through temporarily Dual-GAT (Temporal Dual-Graph Attention Network), dynamically encoding capital flow paths and evolution features over time. Subsequently, in order to identify the structural fraud behavior together produced by colluding subjects, a meta-path subgraph reasoning module based on contrastive learning and hierarchical graph sampling is proposed to enhance the discrimination capability of cross-industry recurring money laundering behavior. Meanwhile, a self-supervised online learning mechanism is adopted for real-time adaptation and continuous optimization to new money laundering strategies. The experimental results show that compared with existing graph neural network methods in cross-industry scenarios, GCRMF improves the performance by more than 17.8% of F1 score and greatly reduces the false positive rate.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
An Integrated Forecasting Prototype for Emergency Department Boarding Time to Support Proactive Operational Decision Making
arXiv:2605.18839v1 Announce Type: new
Abstract: Overcrowding in emergency departments (ED) remains a persistent operational challenge worldwide, causing delays in care delivery and downstream congestion. ED boarding time, defined as the duration admitted patients remain in the ED while awaiting inpatient bed placement, is a key indicator of this congestion. Predicting ED boarding time in advance enables proactive operational decision making before congestion escalates. We developed and evaluated a multi-horizon time series forecasting framework to predict ED boarding time at 6, 8, 10, 12, and 24-hour horizons. Real-world data from a university-affiliated urban hospital in the United States were utilized and integrated with external contextual data sources, including weather, holidays, and major local events. Decomposition-based Linear (DLinear) and Normalization-based Linear (NLinear) time series forecasting deep learning models showed superior performance across multiple horizons. Models were also evaluated under extreme congestion scenarios characterized by elevated boarding times. In addition, a Machine Learning Operations (MLOps) web application prototype was developed to support translation of the forecasting framework into practice through integrated data ingestion, forecast visualization, experimentation, and retraining.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Automated Big Data Quality Assessment using Knowledge Graph Embeddings
arXiv:2605.18833v1 Announce Type: new
Abstract: Automated data quality assessment is crucial for managing big data, but existing solutions face challenges in achieving accurate context-aware assessment. This paper presents a novel knowledge-based approach to enhance automated data quality assessment. Our approach utilizes knowledge graph embeddings to predict missing edges between the input dataset's context representation and the relevant quality rules and dimensions within a knowledge graph representing contextual data characteristics and the required quality assessment operations. We surpass conventional practices by integrating diverse representations within the knowledge graph, drawing insights from contextual information from a thorough literature investigation. This integration allows us to develop a comprehensive and context-specific data quality assessment plan tailored to each context. Leveraging the knowledge graph improves our understanding of the input dataset's context, overcoming the limitations of traditional methods that rely solely on strict matching and overlook contextual characteristics. By injecting numerical edge attributes, we assign corresponding weights to each predicted quality measurement, providing a comprehensive data quality assessment plan for the input dataset.
To evaluate our approach, we leverage AmpliGraph, a framework developed and benchmarked by AccentureLabs. The evaluation involves employing a real-world radiation sensors dataset provided by the Lebanese Atomic Energy Commission (LAEC-CNRS). The results obtained from this evaluation demonstrate the capability of our solution to generate a comprehensive data quality assessment plan for the given input dataset.
Fonte: arXiv cs.LG
MLOps/Systems • Score 92
Not All Tokens Are Worth Caching: Learning Semantic-Aware Eviction for LLM Prefix Caches
arXiv:2605.18825v1 Announce Type: new
Abstract: Prefix caching is a key optimization in Large Language Model (LLM) serving, reusing attention Key-Value (KV) states across requests with shared prompt prefixes to reduce expensive prefill computation. However, its benefit depends critically on the eviction policy as GPU memory is scarce, and existing policies such as LRU largely treat cached blocks uniformly. This view ignores a fundamental property of LLM prompts: not all tokens are equally worth caching. We show that different token types within a prompt, including system prompts, user queries, tool outputs, model responses, and chain-of-thought reasoning, exhibit up to 756x variation in reuse rates, yet no existing eviction policy exploits this signal. In this paper, we present SAECache (Semantic-Adaptive Eviction for prefix caches), a semantic-adaptive prefix cache eviction policy that addresses this gap through three innovations: (1) a multi-queue architecture that routes KV blocks to task-specific queues with tailored priority metrics, capturing both session reuse in multi-turn requests and structural reuse in templated single-turn requests; (2) a semantic-aware token weighting mechanism that learns the reuse value of different token types online through eviction feedback; and (3) a fully adaptive online learning schema for all parameter updates, including log-normal timing parameters, position decay power, queue weights, and meta-parameters, which eliminates manual tuning and enables automatic adaptation to deployment-specific workload characteristics. Through extensive evaluation across heterogeneous workloads, we demonstrate that SAECache achieves 1.4x-2.7x TTFT improvement over production-style baselines, while fixed-parameter alternatives can degrade by up to 2.7x under workload mismatch -- a failure mode our adaptive approach avoids entirely.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
D-PACE: Dynamic Position-Aware Cross-Entropy for Parallel Speculative Drafting
arXiv:2605.18810v1 Announce Type: new
Abstract: Speculative decoding accelerates LLM inference by having a small drafter propose tokens that a larger target model verifies in parallel. Recent diffusion-based parallel drafters such as DFlash predict the full B-token block in one forward pass, enabling deeper drafters and longer accepted blocks. However, existing multi-token drafter objectives often use fixed position-dependent weighting schedules, such as head-dependent weights or block-position decays, which do not adapt as the positions limiting acceptance change during training. To address this, we derive per-position training weights from a differentiable surrogate of expected accepted draft length, matching the weight of each position to its log-probability gradient contribution. The resulting loss, D-PACE (Dynamic Position-Aware Cross-Entropy), shifts training signal toward positions that currently limit acceptance as the drafter improves. Across six benchmarks, two Qwen3-4B draft depths, two decoding temperatures, and two additional target models, D-PACE consistently improves both wall-clock speedup and average emitted length, with 2.3\% measured training-time overhead and no changes to the drafter architecture or inference procedure.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
UCCI: Calibrated Uncertainty for Cost-Optimal LLM Cascade Routing
arXiv:2605.18796v1 Announce Type: new
Abstract: LLM cascades and model routing promise lower inference cost by sending easy queries to a small model and escalating hard ones to a large model, but most deployed routers use uncalibrated confidence scores and require per-workload threshold tuning. We present UCCI, a calibration-first router that maps token-level margin uncertainty to a per-query error probability via isotonic regression and selects the escalation threshold by constrained cost minimization. Under three explicit assumptions, threshold policies on the calibrated score are cost-optimal, and isotonic calibration achieves O(n^{-1/3}) sample complexity for expected calibration error (ECE). On a production named entity recognition workload of 75,000 queries served by 4B and 12B instruction-tuned LLMs on H100 GPUs, UCCI cuts inference cost by 31% (95% CI: [27%, 35%]) at micro-F1 = 0.91 while reducing ECE from 0.12 to 0.03. At the same operating point, UCCI beats entropy thresholding, split-conformal routing, and a FrugalGPT-style learned threshold. All cascade results use end-to-end routing on actual model outputs and measured H100 latency, not simulated routing from global accuracies or nominal API prices.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Robust Basis Spline Decoupling for the Compression of Transformer Models
arXiv:2605.18794v1 Announce Type: new
Abstract: Decoupling is a powerful modeling paradigm for representing multivariate functions as compositions of linear transformations and univariate nonlinear functions. A single-layer decoupling can be viewed as a fully connected neural network with a single hidden layer and flexible activation functions, providing a direct link with neural networks. Because of this, the use of decoupling methods has gained increasing attention in neural network domains, particularly compression, since it enables structured approximations with reduced parameter complexity. Existing tensor-based decoupling methods typically rely on polynomial or piecewise-linear parameterizations of the internal nonlinear functions, which can suffer from numerical instability or limited expressiveness. In this work, we introduce a B-spline-based decoupling framework that generalizes these existing approaches. By exploiting the local support and flexible smoothness control of B-splines, the proposed formulation yields a more numerically stable and expressive representation. We derive a constrained coupled matrix-tensor factorization and propose a robust alternating least-squares algorithm, called R-CMTF-BSD, incorporating normalization and Tikhonov regularization. The proposed method is validated through experiments on synthetic data and transformer model compression. Results on the Vision and Swin Transformer architectures demonstrate that B-spline decoupling enables substantial parameter reduction while maintaining competitive accuracy, making the R-CMTF-BSD algorithm a promising tool for structured neural network compression.
Fonte: arXiv cs.LG
RL • Score 85
Projecting Latent RL Actions: Towards Generalizable and Scalable Graph Combinatorial Optimization
arXiv:2605.19721v1 Announce Type: new
Abstract: Graph combinatorial optimization (GCO) has attracted growing interest, as many NP-hard problems naturally admit graph formulations, yet their combinatorial explosion renders exact methods computationally intractable. Recent advances in Reinforcement Learning (RL) combined with Graph Neural Networks (GNNs) have significantly improved learning-based GCO solvers. However, existing approaches face limitations in both generalization across diverse graph instances and computational scalability as action spaces grow. To address both challenges, we introduce projection agents, a novel RL-GCO approach that operates directly in a continuous GNN-based action embedding space, predicting a desired latent action in a single forward pass and subsequently decoding it into a valid discrete action. Additionally, we enable fair comparison across RL methods through a shared embedding space for both observations and actions. Across diverse benchmarks, our approach achieves up to 16.2x faster inference and up to 40% better generalization than existing solutions using only simple nearest-neighbor decoding, while opening the door to strong RL performance in super-linear decision spaces with multiple interdependent variables. Finally, we release LaGCO-RL, a Python library that automates latent action-space construction and supports existing RL-GCO solutions, promoting reproducibility and adaptation to new GCO benchmarks.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
HELLoRA: Hot Experts Layer-Level Low-Rank Adaptation for Mixture-of-Experts Models
arXiv:2605.18795v1 Announce Type: new
Abstract: Low-Rank Adaptation (LoRA) dominates parameter-efficient fine-tuning of large language models, yet most variants target dense architectures. Mixture-of-Experts (MoE) models scale parameters at near-constant per-token compute, and their sparse activation patterns create untapped opportunities for more efficient adaptation. We propose Hot-Experts Layer-level Low-Rank Adaptation (HELLoRA), which attaches LoRA modules only to the most frequently activated experts at each layer. This simple mechanism reduces trainable parameters and adapter-induced FLOPs while improving downstream performance, an effect we attribute to a form of structured regularization that preserves pretrained expert specialization. To stress-test HELLoRA under extreme parameter budgets, we further compose it with LoRI to form HELLoRI, which freezes the up-projection and sparsifies the down-projection. Across three MoE backbones, namely OlMoE-1B-7B, Mixtral-8x7B, and DeepSeekMoE, and three task families covering mathematical reasoning, code generation, and safety alignment, HELLoRA consistently outperforms strong PEFT baselines. Relative to vanilla LoRA on OlMoE, HELLoRA uses 15.7% of the trainable parameters, reduces adapter FLOPs by 38.7%, achieves 1.9x the training throughput, and improves accuracy by 9.2%. On DeepSeekMoE, HELLoRA outperforms LoRA while using only 23.2% of its trainable parameters. These results demonstrate that activation-aware adapter placement is an effective and practical route to scaling PEFT for MoE language models.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
DynaTrain: Fast Online Parallelism Switching for Elastic LLM Training
arXiv:2605.18815v1 Announce Type: new
Abstract: Modern large language model (LLM) training is inherently dynamic: resource fluctuations, RLHF phase shifts, and cluster elasticity continually reshape the optimal parallelism layout, posing a significant challenge to existing training frameworks built around a static execution model. We present DynaTrain, a distributed training system for sub-second, online reconfiguration across arbitrary multi-dimensional parallelism. At its core, we propose a Virtual Parameter Space (VPS) abstraction that unifies all distributed training states under one logical coordinate space, turning any parallelism configuration into a deterministic mapping and collapsing complex transition into manageable geometric intersections. On top of VPS, a state routing-and-transition layer executes rank-local transfers under a memory-aware, deadlock-free schedule, and an Elastic Device Manager overlaps new-world construction with ongoing training to mask topology-change cost. On dense and MoE models up to 235B parameters, DynaTrain reconfigures a 70B dense model in under 2s and a 235B MoE model in 4.36s, outperforming state-of-the-art checkpoint-based and elastic systems by up to three orders of magnitude while preserving correctness.
Fonte: arXiv cs.LG
RecSys • Score 85
SimGym: A Framework for A/B Test Simulation in E-Commerce with Traffic-Grounded VLM Agents
arXiv:2605.19219v1 Announce Type: new
Abstract: A/B testing remains the gold standard for evaluating modifications to e-commerce storefronts, yet it diverts traffic, requires weeks to reach statistical significance, and risks degrading user experience. We present SimGym, a framework for simulating A/B tests on e-commerce storefronts using vision-language model (VLM) agents operating in a live browser. The framework comprises three key components: (a) a traffic-grounded persona generation pipeline that derives per-shop buyer archetypes and intents from production clickstream data; (b) a live-browser agent architecture that combines multimodal perception over visual and browser-structured observations with episodic memory and guardrails to conduct coherent shopping sessions across control and treatment storefronts; and (c) an evaluation protocol that compares simulated outcome shifts with observed shifts in real buyer behavior. We validate SimGym on A/B tests of visually driven UI theme changes from a major e-commerce platform across diverse storefronts and product categories. Empirical results show that SimGym agents achieve strong agreement with observed outcome shifts, attaining 77% directional alignment with add-to-cart shifts observed across interface variants in real-buyer traffic. It reduces experimental cycles from weeks to under an hour, enabling rapid experimentation without exposing real buyers to candidate variants.
Fonte: arXiv cs.AI
Vision • Score 85
INAR-VL: Input-Aware Routing for Edge-Cloud Vision-Language Inference
arXiv:2605.18853v1 Announce Type: new
Abstract: Edge deployment of Vision-Language Models (VLMs) faces a tradeoff between latency and accuracy: cloud execution provides high-quality predictions but incurs communication delay and energy cost, while edge-only execution is faster but less accurate due to limited model capacity. This trade-off is further complicated by heterogeneity in image quality and reasoning complexity, making static placement suboptimal. We present INAR-VL, a lightweight edge-cloud routing system for multimodal inference in a two-tier deployment. INAR-VL maintains complementary VLMs across edge and cloud and uses lightweight image and text complexity signals to guide routing and model selection, executing simple queries locally while offloading complex ones when beneficial. Evaluation on visual question answering shows that INAR-VL executes 36% of requests on the edge, reduces latency by 24%, lowers energy by 26%, and preserves 97% of cloud-level accuracy.
Fonte: arXiv cs.LG
RL • Score 85
PROWL: Prioritized Regret-Driven Optimization for World Model Learning
arXiv:2605.18803v1 Announce Type: new
Abstract: Modern action-conditioned video world models achieve strong short-horizon visual realism, yet remain unreliable on rare, interaction-critical transitions that dominate downstream planning and policy performance. Because passive demonstration data systematically under-samples these high-impact regimes, improving robustness requires actively eliciting model failures rather than relying on their natural occurrence. We introduce a KL-constrained adversarial curriculum in which a policy is trained to expose high-error trajectories of a diffusion-based world model while remaining close to the behavior distribution. The world model is continuously fine-tuned on these adversarially discovered trajectories, yielding an adversarial training loop that converts rare failures into a stable, near-distribution training signal without drifting into out-of-distribution exploitation. To maintain pressure on unresolved weaknesses as the model improves, we propose a Prioritized Adversarial Trajectory (PAT) buffer that re-ranks trajectories based on prediction error, action fidelity, and learning progress, focusing training on unresolved failure modes rather than repeatedly revisiting solved cases. We implement our approach in the MineRL framework and evaluate it on held-out out-of-distribution trajectories; PROWL improves robustness over models trained on passive data alone, reveals reward-hacking behaviors under weak behavioral constraints, and demonstrates that effective adversarial world-model training critically depends on balancing exploratory failure discovery with explicit behavioral regularization. Our results suggest that scalable world models benefit not only from larger datasets, but also from selectively generating informative training data.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Formal Skill: Programmable Runtime Skills for Efficient and Accurate LLM Agents
arXiv:2605.19604v1 Announce Type: new
Abstract: Large Language Model (LLM) agents increasingly act inside real workspaces, where tools and skills determine whether model reasoning becomes reliable action. Existing skills remain largely informal: Markdown skills and instruction packs encode procedures as long natural-language documents, while function calling, Model Context Protocol (MCP) servers, and framework tools structure individual actions but usually leave workflow state, policy enforcement, and completion discipline outside the skill itself. We introduce Formal Skill, a runtime-native abstraction that represents reusable capability with JSON metadata and action schemas, reliable Python executors, hook-governed control logic, Formal Skill routing, and skill-local runtime state. By moving reusable procedure from repeated prompt text into executable state machines and hook policies, Formal Skill gives agents a token-efficient and enforceable control surface. We implement the abstraction in FairyClaw, an open-source event-driven runtime for executable, observable, and composable Formal Skills. On Harness-Bench, FairyClaw obtains highly competitive average scores while using substantially fewer tokens, with especially strong results on tasks that expose the role of Formal Skill.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Towards Multi-Model LLM Schedulers: Empirical Insights into Offloading and Preemption
arXiv:2605.19593v1 Announce Type: new
Abstract: Modern deployments of Large Language Models (LLMs) increasingly require serving multiple models with diverse architectures, sizes, and specialization on shared, heterogeneous hardware. This setting introduces new challenges for resource allocation, dispatching, and scheduling, particularly under GPU memory constraints where partial CPU-GPU offloading and preemption become necessary. While existing systems primarily optimize throughput for a single model, comparatively little work addresses multi-model scheduling under these conditions. In this paper, we present an empirical study of how different LLMs behave across hardware platforms, focusing on the performance implications of layer offloading and preemption. We show that offloading leads to strongly non-linear and model-dependent degradation in decode throughput, with smaller models exhibiting sharper sensitivity to reduced GPU residency. We further demonstrate that preemption incurs substantial overhead, largely dominated by model state reload rather than key-value cache transfer, and that this cost varies significantly across models and hardware platforms. Additionally, we highlight the role of sequence length and interconnect bandwidth in amplifying data movement and execution inefficiencies. Based on these findings, we identify a set of key features that future schedulers must consider, including model-specific offloading sensitivity, workload characteristics, and the cost structure of preemption and data transfer. These insights provide guidance for the design of next-generation LLM serving systems capable of efficiently managing heterogeneous, multi-model workloads with hybrid CPU-GPU execution.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Operationalizing Document AI: A Microservice Architecture for OCR and LLM Pipelines in Production
arXiv:2605.18818v1 Announce Type: new
Abstract: Academic research tends to focus on new models for document understanding creating a wide gap in the literature between model definition and running models at production scale. To close that gap, we present a microservice architecture that encapsulates pipelines of multiple models for classification, optical character recognition (OCR), and large language model structured field extraction as well as our experience running this pipeline on thousands of multi-page documents per hour. We describe our primary design decisions, including a hybrid classification, separation of GPU-bound inference from CPU-bound orchestration, use of asynchronous processing for the many IO-bound operations in the pipeline, and an independent, horizontal scaling strategy. Using batch profiling, we identified two surprising qualitative findings that shape production deployments: OCR, not language-model parsing, dominates end-to-end latency, and the system saturates at a concurrency determined by shared GPU-inference capacity rather than worker count. Our goal is to provide practitioners with concrete architectural patterns for building document understanding systems that work beyond the benchmark; effectively operationalizing models in production.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
The Thermodynamic Costs of Simple Linear Regression
arXiv:2605.19195v1 Announce Type: cross
Abstract: The construction of models from data is a significant contributor to the energetic costs of computation. Because of this, understanding how foundational thermodynamic bounds apply to modeling algorithms will be increasingly important. Here, we study the thermodynamic costs of a basic and fundamental modeling algorithm: simple linear regression. Following Landauer, we approximate the thermodynamic lower bound on irreversibly performing both exact linear regression and linear regression via stochastic gradient descent as implemented on floating-point numbers. From this, we derive energycost aware scaling laws for the optimal dataset size for training a linear regression model given a generalization error dependent demand for inference. Additionally, we discuss a method to lower bound the entropy production from the mismatch cost for algorithms with continuous input variables.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Embedding by Elicitation: Dynamic Representations for Bayesian Optimization of System Prompts
arXiv:2605.19093v1 Announce Type: new
Abstract: System prompts are a central control mechanism in modern AI systems, shaping behavior across conversations, tasks, and user populations. Yet they are difficult to tune when feedback is available only as aggregate metrics rather than per-example labels, failures, or critiques. We study this aggregate feedback setting as sample-constrained black-box optimization over discrete, variable-length text. We introduce ReElicit, a Bayesian optimization framework based on \emph{embedding by elicitation}. Given a task description, previously evaluated prompts, and scalar scores, an LLM elicits a compact, interpretable feature space and maps prompts into it. Leveraging a probabilistic Gaussian process surrogate, an acquisition function then selects target feature vectors, which the LLM realizes and refines into deployable system prompts. Re-eliciting the feature space as new evaluations arrive lets the representation adapt to the observed prompt-score history. We evaluate the setting using offline benchmark accuracy as a controlled aggregate proxy: the optimizer observes one scalar score per prompt and no per-example labels, errors, or critiques. Across ten system prompt optimization tasks with a 30 total evaluation budget, ReElicit achieves the strongest aggregate performance profile among representative aggregate-only prompt-optimization baselines. These results suggest that LLMs can serve as adaptive semantic representation builders, not only prompt generators, for Bayesian optimization over natural-language artifacts.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Towards Data-Efficient Video Pre-training with Frozen Image Foundation Models
arXiv:2605.19137v1 Announce Type: new
Abstract: Video foundation models achieve strong performance across many video understanding tasks, but typically require large-scale pre-training on massive video datasets, resulting in substantial data and compute costs. In contrast, modern image foundation models already provide powerful spatial representations. This raises an important question: can competitive video models be built by reusing these spatial representations and pre-training only for temporal reasoning? We take initial steps toward exploring a lightweight training paradigm that freezes a pre-trained image foundation model and trains only a recurrent temporal module to process streaming video. By reusing an image foundation model as a spatial encoder, this approach could significantly reduce the amount of video data and compute required compared to end-to-end video pre-training. In this work, we explore the feasibility of this approach before investing in computing for video pre-training. Our empirical findings across multiple video understanding tasks suggest that strong temporal performance can emerge without large-scale video pre-training, motivating future work on recurrent video foundation models obtained by pre-training a temporal module on top of a frozen image foundation model. Code: https://github.com/tue-mps/towards-video-image-frozen .
Fonte: arXiv cs.CV
Privacy/Security/Fairness • Score 85
Federated Learning with Nonvacuous Generalisation Bounds
arXiv:2310.11203v2 Announce Type: replace-cross
Abstract: We introduce a novel strategy to train randomised predictors in federated learning, where each node of the network aims at preserving its privacy by releasing a local predictor but keeping secret its training dataset with respect to the other nodes.
We then build a global randomised predictor which inherits the properties of the local private predictors in the sense of a PAC-Bayesian generalisation bound. We consider the synchronous case where all nodes share the same training objective (derived from a generalisation bound), and the heterogenous and homogenous cases where each node may have its own personalised training objective. We show through a series of numerical experiments that our approach achieves a comparable predictive performance to that of the batch approach where all datasets are shared across nodes. Moreover the predictors are supported by numerically nonvacuous generalisation bounds while preserving privacy for each node. We explicitly compute the increment on predictive performance and generalisation bounds for our two federated settings, highlighting the price to pay to preserve privacy.
Fonte: arXiv stat.ML
Theory/Optimization • Score 85
Theory-optimal Quantization Based on Flatness
arXiv:2605.18800v1 Announce Type: new
Abstract: Post-training quantization has emerged as a widely adopted technique for compressing and accelerating the inference of Large Language Models (LLMs). The primary challenges in LLMs quantization stem from activation outliers, which significantly degrade model performance especially at lower bit precision. While recent approaches attempt to mitigate outliers through linear transformations across feature dimensions, our analysis reveals that the transformed weights and activations still exhibit persistent outlier patterns with concentrated magnitude distributions. In this paper, we first model the mathematical relationship between quantization error and outliers, and then introduce a new metric Flatness to quantify the distribution of outliers. Based on this, we derive the theoretical optimal solution with respect to Flatness. Building on these insights, we propose Bidirectional Diagonal Quantization (BDQ), a novel post-training quantization framework that effectively disperses outlier patterns through optimized matrix transformations. BDQ strategically distributes outlier magnitudes across matrix dimensions via learned diagonal operations. Extensive experiments demonstrate that BDQ establishes a new quantization benchmark. It achieves less than 1\% accuracy drop in W4A4 quantization on the LLaMA-3-8B model. In the more challenging W2A4KV16 experiment, compared to state-of-the-art approaches, BDQ reduces the performance gap by 39.1\% on the DeepSeek-R1-Distill-LLaMA-70B model.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Factor Augmented High-Dimensional SGD
arXiv:2605.19291v1 Announce Type: new
Abstract: Stochastic gradient descent (SGD) is a fundamental optimization algorithm widely used in modern machine learning. In this paper, we propose Factor-Augmented SGD (FSGD), a new optimization method that leverages latent factor representations in high-dimensional learning tasks. Unlike standard two-stage dimension reduction approaches that rely on offline representation learning and full data storage, a key novelty of FSGD is that it operates purely on streaming data, making it scalable to large-scale and high-dimensional problems. Furthermore, we establish the first theoretical framework that explicitly incorporates latent factor estimation error into the analysis of SGD, and provide moment convergence in $\ell^s$ norm under decaying step sizes and mini-batch updates. Our results provide a new foundation for employing SGD reliably and scalably in high-dimensional machine learning systems.
Fonte: arXiv stat.ML
RL • Score 85
GoLongRL: Capability-Oriented Long Context Reinforcement Learning with Multitask Alignment
arXiv:2605.19577v1 Announce Type: new
Abstract: We present GoLongRL, a fully open-source, capability-oriented post-training recipe for long-context reinforcement learning with verifiable rewards (RLVR). Existing long-context RL methods often treat data construction as a matter of designing increasingly complex retrieval paths, leading to homogeneous task coverage and reward formulations that inadequately reflect practical long-context requirements. Our work offers two contributions. (1) Capability-oriented data construction with full open release. We openly release a dataset of 23K RLVR samples, the complete construction pipeline, and all training code. Guided by a taxonomy of long-context capabilities, the dataset spans 9 task types, each paired with its natural evaluation metric. It comprises curated open-source samples from established corpora and synthetic samples whose QA pairs are generated from real source documents such as books, academic papers, and multi-turn dialogues. Under the same vanilla GRPO setup, our dataset alone outperforms the closed-source QwenLong-L1.5 dataset. Moreover, our Qwen3-30B-A3B model trained on this data delivers long-context performance comparable to DeepSeek-R1-0528 and Qwen3-235B-A22B-Thinking-2507, suggesting that broader coverage and greater reward diversity substantially benefit long-context capability improvement. (2) TMN-Reweight for heterogeneous multitask optimization. To address optimization challenges from heterogeneous rewards, we propose TMN-Reweight, which combines task-level mean normalization for cross-task reward scale alignment with difficulty-adaptive weighting for more reliable advantage estimation. TMN-Reweight further improves average performance over vanilla GRPO, with general capabilities preserved or improved across reported evaluations.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
EngiAI: A Multi-Agent Framework and Benchmark Suite for LLM-Driven Engineering Design
arXiv:2605.19743v1 Announce Type: new
Abstract: Large Language Model (LLM) agents are increasingly applied to engineering design tasks, yet existing evaluation frameworks do not adequately address multi-agent systems that combine simulation, retrieval, and manufacturing preparation. We introduce a benchmark suite with three evaluation dimensions: (1) a workflow benchmark with seven prompt styles targeting distinct cognitive demands-including direct tool use, semantic disambiguation, conditional branching, and working-memory tasks; (2) a Retrieval-Augmented Generation (RAG) benchmark with gated scoring isolating retrieval contributions to parameter selection; and (3) an High Performance Computing (HPC) benchmark evaluating end-to-end ML training orchestration on a SLURM cluster. Alongside the benchmark we present EngiAI, a Multi-Agent System (MAS) reference implementation built on LangGraph that operationalizes the benchmark by coordinating seven specialized agents through a supervisor architecture, unifying topology optimization, document retrieval, HPC job orchestration, and 3D printer control. Across four LLM backends and two EngiBench problems, proprietary models achieve 96-97% average task completion on Beams2D, while open-source 4B-parameter models reach 55-78%, with clear generational improvement. Conditional branching proves most challenging, with task completion dropping to 20-53% for the conditional style on Photonics2D. RAG gating confirms near-perfect retrieval-augmented scores ($\approx 1.0$) versus near-zero without retrieval, validating the evaluation design. On HPC orchestration, one model completes all pipeline steps in 100% of runs while another drops to 50%, revealing that multi-step instruction following degrades over long-running workflows.
Fonte: arXiv cs.AI
Multimodal • Score 85
VCR: Learning Valid Contextual Representation for Incomplete Wearable Signals
arXiv:2605.18837v1 Announce Type: new
Abstract: Wearable devices enable continuous health monitoring from multimodal signals, but real-world deployment is hindered by limited labeled data and pervasive sensor incompleteness. While large-scale self-supervised pretraining reduces label dependence, most existing methods assume full modality availability. Current approaches for handling modality missingness often reconstruct entire absent signals, which can encourage hallucinating modality-specific details that are not inferable from the observed sensor signals and degrade robustness. We propose VCR, a self-supervised framework that learns to extract valid representations robust to modality missingness. VCR employs an orthogonal tokenizer to enforce strict orthogonal disentanglement by rectifying latent manifolds and applying a geometric projection, separating each modality into shared semantics and modality-specific residuals. This design preserves complete information integrity while serving as a structural foundation for robust learning under modality missingness. The resulting tokens are processed by a missing-aware mixture-of-experts backbone that adapts to varying patterns of modality availability. By constraining the objective to reconstruct only the shared components of missing modalities, VCR effectively mitigates hallucinations of non-inferable modality-specific details. Across multiple health monitoring tasks, VCR consistently improves performance and robustness under full, single-missing, and multiple-missing modality settings compared with strong supervised and self-supervised baselines.
Fonte: arXiv cs.LG
Vision • Score 85
Rotation-Aligned Key Channel Pruning for Efficient Vision-Language Model Inference
arXiv:2605.19218v1 Announce Type: new
Abstract: Vision-Language Models suffer severe KV cache pressure at inference, as a single image often encodes into thousands of tokens. Most existing methods exploit token sparsity through token pruning, but permanently discarding visual content causes substantial degradation on fine-grained perception tasks. This motivates a complementary axis, feature sparsity: under a fixed KV cache budget, compressing the channel dimension preserves more visual tokens at the same memory cost. Prior Key channel pruning methods, however, face a structural trade-off: token-wise channel pruning is expressive but unstructured and slow, while head-wise approach is hardware-friendly but less robust. We resolve this with RotateK, a rotation-based structured Key channel pruning framework. RotateK applies an online PCA-based rotation that aligns token-dependent channel importance into a shared low-dimensional subspace, enabling accurate pruning under lightweight head-wise masks; a fused Triton attention kernel operates directly on sparse-channel Keys for efficient decoding. Experiments on two representative VLM backbones show that RotateK consistently outperforms prior Key channel pruning in both accuracy and decoding latency, while joint token-channel pruning improves over token-only baselines at matched KV cache budgets.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Structuring Open-Ended NAS: Semi-Automated Design Knowledge Structuring with LLMs for Efficient Neural Architecture Search
arXiv:2605.19247v1 Announce Type: new
Abstract: Current neural architecture search (NAS) methods are often limited by their predefined, restrictive search spaces. While recent large language model (LLM)-assisted NAS methods enable open-ended search spaces, they often suffer from inefficient exploration due to biased or low-quality design ideas. To address these issues, we propose to semi-automatically structure model design knowledge to guide the search process. Our approach first defines a high-level structural template of architectural attributes. An LLM then populates this template by analyzing papers, creating a rich and diverse search space that embodies this structured design knowledge. To efficiently explore this vast space, we introduce FairNAD, using a multi-type mutation that enables broad exploration through mutation with fair idea sampling, Pareto-aware mutation, LLM-driven iterative mutation, and a fine-grained feedback loop. We demonstrate the effectiveness of FairNAD in discovering high-performing architectures that yield 0.84, 2.17, and 2.35 points improvement on CIFAR-10, CIFAR-100, and ImageNet16-120, respectively, compared to current state-of-the-art methods.
Fonte: arXiv cs.CV
Evaluation/Benchmarks • Score 85
Augmenting Human Evaluation with LLM Judges: How Many Human Reviews Do You Need?
arXiv:2605.16354v1 Announce Type: new
Abstract: Large language models (LLMs) are increasingly used as automated evaluators of AI systems, including in high-stakes applications. In this role, LLMs are used to generate judgments about the quality, appropriateness, or even safety of model outputs. This approach is motivated by practical constraints. Expert human ratings are costly and difficult to scale, whereas LLM ratings can be produced quickly at low cost. However, current approaches to deploying LLM evaluators are ad hoc, typically limited to reporting agreement metrics between human and LLM judges as a justification for substitution of human ratings, and lack a formal basis for study design. This paper (1) shifts the role of the LLM judge from substitutive to auxiliary, and (2) formulates the LLM-as-a-judge paradigm as one of augmenting human evaluation through a two-stage sampling design, where LLM evaluations are measured for all observations at the first stage and human ratings are partially observed for a subsample at the second stage. We propose to use a doubly robust estimator from the missing data literature, which takes advantage of the robustness property against the prediction model, since the missingness model is known by design. Using the asymptotic variance of this estimator, we propose how sample sizes of human and LLM ratings can be determined to achieve a targeted level of power. We also show that a study can be efficiently designed by allocating more human ratings for types of evaluations where the predictability of LLM ratings is not high. To the best of our knowledge, there is very little guidance on how much human oversight should be retained when validating benchmarks.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
SWoMo: Neuro-Symbolic World Model for Cataract Surgery Simulation
arXiv:2605.16530v1 Announce Type: new
Abstract: Realistic surgical simulation plays a crucial role in training novice surgeons and in the development of autonomous agents. World models can scale such simulation environments to realistic and diverse procedures by predicting future patient states conditioned on current observations and surgical actions. However, current state-of-the-art approaches often fail to satisfy key criteria required for clinical applicability, including visual realism, physically grounded interactions, and the ability to simulate scenarios beyond the training distribution. Hence, we introduce SWoMo, a neuro-symbolic world model for cataract surgery simulation that decouples motion generation from visual realism. The symbolic component, consisting of a rule-based simulator and scene graph representations, models motion dynamics and tool-tissue interactions, while a diffusion model produces realistic visual appearance, including textures and tissue deformations. We propose an inverse pairing strategy that reconstructs real surgical videos in the simulator to obtain paired simulated and real videos, which are then used to train our video diffusion model for the reverse objective of sim-to-real translation. Our experiments show both qualitative and quantitative improvements over prior work. We demonstrate that our simulator further satisfies the key criteria, including generalisation to unseen interaction geometries, improvements in downstream phase detection, and unsupervised video style transfer. The code, data, and model weights are available at: https://ssharvienkumar.github.io/SWoMo/
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
Nonlinear Bipolar Compensation: Handling Outliers in Post-Training Quantization
arXiv:2605.16423v1 Announce Type: new
Abstract: Network quantization has emerged as one of the most practical model compression techniques, which significantly reduces a model's memory and compute consumption by mapping floating-point numbers to low-bit representations. However, existing quantization methods typically suffer from the speed-accuracy tradeoff and limited generalization. To address these issues, recent compensation-based methods offer an efficient yet general solution by introducing additional lightweight linear layers into the quantized network. However, the accuracy of these methods suffers from their limited compensation capability and high sensitivity to outliers. In this paper, we propose Nonlinear Bipolar Compensation (NBC), a post-training quantization approach that introduces nonlinear compensation to reduce the effect of outliers. We further design Bipolar Logarithmic Transformation (BLT), which compresses outliers by mapping both the quantized input and the quantization error into a transformed space. A simple linear layer is then applied for compensation in the transformed space, preserving the efficiency of our method. Extensive experiments across various tasks, models, and quantization methods confirm the effectiveness, efficiency, robustness, and generality of our NBC approach.
Fonte: arXiv cs.CV
Vision • Score 85
KVCapsule: Efficient Sequential KV Cache Compression for Vision-Language Models with Asymmetric Redundancy
arXiv:2605.16439v1 Announce Type: new
Abstract: Vision-Language Models (VLMs) have emerged as a critical and fast-growing extension of Large Language Models (LLMs) that enable multimodal reasoning through both text and image inputs. Although VLMs enrich the capabilities of language models, they also inherit and amplify key computational bottlenecks: the memory overhead caused by the large key-value (KV) cache during autoregressive decoding. This challenge is particularly severe in VLMs, where images produce longer token sequences and denser feature representations compared to text. Moreover, the spatial and information-rich nature of vision tokens introduces structured attention patterns that make many LLM-oriented KV cache compression techniques ineffective when applied directly to VLMs.
In this work, we conduct a detailed empirical analysis of the behavior of vision tokens, highlighting the critical differences from purely text-based models. Based on these insights, we propose KVCapsule, a novel KV cache compression framework for vision tokens. KVCapsule keeps the pretrained VLM backbone frozen, requires no modification to the attention computation modules, and can be integrated into existing VLMs through lightweight compression and reconstruction components. We evaluate KVCapsule on multiple VLMs and benchmark tasks, demonstrating up to 2x improvement in TPS and 2.4x reduction in KV cache memory at a 60% compression ratio, with negligible degradation in accuracy or response quality. Our findings offer practical pathways to scale VLM inference under constrained memory budgets and inspire further research into structure-aware cache compression for multimodal models.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
Your SaaS Is an Insurance Product: A Modeling Framework
arXiv:2605.16699v1 Announce Type: cross
Abstract: Capped-usage SaaS products -- LLM subscriptions such as Claude Code and ChatGPT, cloud platforms such as Vercel and Cloudflare Workers, corporate benefit platforms, identity-verification services with liability transfer -- share a structural signature with insurance products: a fixed premium decoupled from realized consumption, stochastic per-user demand with heavy-tailed severity, a non-fungible cap that resets on a fixed schedule, and a portfolio-level exposure that requires reserve adequacy under tail risk. We argue that this is not an analogy. It is the same operational problem actuarial science has been tooled for decades to address, restated with new dependent variables (tokens, bandwidth bytes, function-invocations, gym check-ins) in place of medical claims. This paper proposes a modeling framework for capped-usage SaaS pricing built from frequency-severity decomposition, premium calculation principles, and Monte Carlo reserve adequacy. We map the framework to publicly observable subscription tiers in two domains (LLM services and cloud platforms), ground it in canonical health-insurance economics (Arrow 1963; Pauly 1968; Manning et al. 1987; Brot-Goldberg et al. 2017), and demonstrate divergence from traditional unit economics through a worked example. The contribution is operational rather than theoretical: not a new theorem, but vocabulary and tools currently absent from cs.LG/stat.ML practice.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Byzantine-Resilient Federated Learning via QUBO-Based Client Selection on Quantum Annealers
arXiv:2605.16438v1 Announce Type: new
Abstract: Federated Learning (FL) trains a global model across decentralized clients while preserving data privacy, but at scale it is vulnerable to malicious updates. Byzantine-resilient aggregation methods such as MultiKrum score gradients against their nearest neighbors and can miss malicious updates that preserve the statistical properties of honest ones. We propose a quantum annealing approach that reformulates client selection as a Quadratic Unconstrained Binary Optimization (QUBO) problem, encoding pairwise distances into a cost function solved by quantum annealers (QA). Unlike MultiKrum's greedy per-client scoring, the QUBO formulation jointly optimizes over all subsets to find the mutually closest group of $m$ clients. At small scale (15 clients), QUBO outperforms MultiKrum on the most challenging Byzantine attacks: e.g., Advanced LIE is detected with 95.11% accuracy versus 81.33% on MNIST and 97.78% versus 75.56% on CIFAR-10. QUBO fares poorly on simpler attacks where MultiKrum excels, so the two methods are complementary. QUBO quality also degrades as the number of clients grows. To address this, we introduce a MultiSignal ensemble that uses a dual-feature routing gate based on Euclidean and cosine Krum score gaps to classify attacks into four regimes and routes evasion attacks to a suspicion-penalized QUBO with agreement voting. At 100 clients on MNIST, MultiSignal achieves 95.3% average detection accuracy versus 91.8% for classical MultiKrum, with the largest gains on Sparse Lie (72.0% to 95.2%, +23.2 points) and Advanced Lie (80.4% to 85.2%, +4.8 points). These results show that QUBO-based quantum annealing with MultiSignal is a principled and scalable defense against the most challenging Byzantine strategies in federated learning.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
AgentKernelArena: Generalization-Aware Benchmarking of GPU Kernel Optimization Agents
arXiv:2605.16819v1 Announce Type: new
Abstract: GPU kernel optimization is increasingly critical for efficient deep learning systems, but writing high-performance kernels still requires substantial low-level expertise. Recent AI coding agents can iteratively read code, invoke compilers and profilers, and refine implementations, yet existing kernel benchmarks evaluate single LLM calls rather than full agent workflows, and none include both kernel-to-kernel optimization and unseen-configuration generalization testing. We present AgentKernelArena, an open-source benchmark for measuring AI coding agents on GPU kernel optimization. The benchmark contains 196 tasks spanning HIP-to-HIP optimization, Triton-to-Triton optimization, and PyTorch-to-HIP translation, and evaluates complete agent workflows in isolated workspaces using gated compilation, correctness, and performance checks, centralized scoring and an unseen-configuration generalization protocol that tests whether optimizations transfer to input configurations the agent never observed. Across production agents including Cursor Agent, Claude Code, and Codex Agent, we find near-perfect compilation and high correctness rates on most task categories, with the strongest configurations achieving mean speedups of up to 6.89x on PyTorch-to-HIP, 6.69x on HIP-to-HIP, and 2.13x on Triton-to-Triton tasks. Our unseen-configuration evaluation shows that HIP-to-HIP and Triton-to-Triton optimizations largely transfer to unseen input shapes, while PyTorch-to-HIP exhibits substantial correctness drops, indicating that agents generating kernels from scratch frequently hardcode shape-specific assumptions. AgentKernelArena is designed as a modular, extensible framework for rigorous evaluation of agentic GPU kernel optimization across agents, tasks, and hardware targets.
Fonte: arXiv cs.CL
Vision • Score 85
Trajectory-Aware Adaptive Inference in Object Detection Models
arXiv:2605.16397v1 Announce Type: new
Abstract: The increasing integration of sensors in autonomous maritime navigation has led to large-scale multimodal datasets, raising challenges in achieving efficient real-time perception. In such systems, object detection and trajectory perception of nearby vessels are tightly coupled, particularly in dynamic environments such as maritime navigation. However, the efficiency of object detection models during inference remains an often-overlooked aspect. To this end, we build upon an existing object detection framework by incorporating GPS trajectory data into the inference process to enable input-adaptive computation. Specifically, we introduce an early-exit mechanism in a YOLOv8-based detector that incorporates motion cues - such as inter-vessel distances. Frames of vessels that are separated by short distances, converging with high speed, are processed using the full model, while only a subset of the network's architecture is activated otherwise. The difficulty degree (or scene complexity) of a frame or set of frames per second is evaluated by leveraging inter-object distance and the rate at which the distance between them decreases. Experimental results demonstrate that this strategy maintains satisfactory detection performance while significantly reducing inference time and computational cost, thus enabling a flexible trade-off between accuracy and efficiency compared to full-model inference.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
Fre-Res: Frequency-Residual Video Token Compression for Efficient Video MLLMs
arXiv:2605.16366v1 Announce Type: new
Abstract: Video MLLMs face a persistent tension between spatial fidelity and temporal coverage: preserving fine-grained visual details requires many spatial tokens, while capturing short-lived events requires dense temporal sampling. We propose \textbf{Fre-Res}, a budget-adaptive dual-track video-token compression framework that separates these two forms of evidence. Fre-Res preserves sparse high-fidelity spatial anchors and represents dense temporal evolution through compact residual-frequency tokens. Specifically, it applies temporal 1D-DCT to inter-frame residual trajectories in vision-latent space, where we observe strong low-frequency concentration. To align frequency-domain dynamics with native visual embeddings, Fre-Res introduces a Spatial-Guided Absorber that injects temporal residual information into spatially corresponding anchor tokens. Across fine-grained short-video and long-video reasoning benchmarks, Fre-Res achieves a favorable accuracy--efficiency trade-off, matching or approaching full-token performance while substantially reducing visual-token length. Extensive ablations further show that temporal-frequency residuals preserve causal transition cues, while spatial anchors remain essential for fine-grained object and layout reasoning.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
Two-Valued Symmetric Circulant Matrices: Applications in Deep Learning
arXiv:2605.16443v1 Announce Type: new
Abstract: Despite the success of deep neural networks in vision, medical diagnosis, and IoT scenarios, their deployment on resource-limited platforms poses serious challenges due to their high storage requirements, computational complexity, and large footprint. In particular, fully connected layers require a large number of weights, making it difficult for edge devices to accommodate them. To overcome these challenges associated with limited platforms, this paper proposes the Two-Valued Symmetric Circulant Matrix (TVSCM), a very sparse architecture that employs just two weights per layer to keep it circulant and symmetric. The extreme form of structured sparse architecture provides negligible storage costs compared to traditional full-weight storage. Instead of hardware and additional stages of other traditional sparse learning techniques, such as low-rank approximation and pruning approaches, this architecture provides an extreme form of sparsity, achieving very minimal storage requirements. The simulation study demonstrates more than 80$\times$ reduction in model parameters, reducing parameters from 623,290 to 7,852 on MNIST and from 24,709 to 942 on the MIT-BIH arrhythmia dataset, while maintaining comparable accuracy from 97.6% to 93.5% on MNIST and from 97.6% to 93.1% on MIT-BIH. Due to its minimal architectural requirements and very low power consumption, this architecture would be ideal for edge computing platforms, tiny-ML platforms, IoMT systems, and battery-powered systems.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
CADS: Conformal Adaptive Decision System for Cost-Efficient Image Classification
arXiv:2605.16401v1 Announce Type: new
Abstract: While high-capacity AI models have advanced state-of-the-art performance, their practical deployment is often hindered by high inference costs, environmental impact, and a "one-size-fits-all" approach that ignores varying sample complexity. In clinical settings for instance, the waste of computational resources on routine cases is a significant barrier to sustainable AI. In this paper, we introduce the Conformal Adaptive Decision System (CADS), a sequential multi-model algorithm designed to optimize resource allocation by efficiently sampling models based on the estimated data complexity. CADS leverages conformal prediction to quantify image uncertainty at runtime. CADS provides a mathematically grounded framework for balancing the cost-accuracy dilemma that dynamically routes samples through a model cascade, ranging from lightweight "Scout" models to high-capacity "Oracle" architectures. Validated on two datasets, CADS demonstrated superior efficiency and accuracy at a computational cost that can be up to 12 times lower than heavy-model inference. By accurately routing samples based on real-time complexity, CADS ensures high diagnostic reliability while drastically reducing the economic and environmental footprint of AI.
Fonte: arXiv cs.CV
Multimodal • Score 85
StrLoRA: Towards Streaming Continual Visual Instruction Tuning for MLLMs
arXiv:2605.16353v1 Announce Type: new
Abstract: Continual Visual Instruction Tuning (CVIT) enables Multimodal Large Language Models to incrementally acquire new abilities. However, existing CVIT methods operate under a restrictive task-incremental setting, where each training phase corresponds to a single, predefined task. This does not reflect real-world conditions, where data arrives as a continuous stream of interleaved and dynamically evolving tasks. To bridge this gap, we introduce Streaming CVIT (StrCVIT), a more general and realistic setting where models learn from a stream of data chunks containing a dynamic mixture of tasks. In StrCVIT, a model must simultaneously acquire new abilities, reinforce recurring abilities, and mitigate forgetting. Existing CVIT methods fail here as they cannot reliably distinguish or adapt to the heterogeneous task samples within each chunk. We therefore propose StrLoRA, a regularized two-stage expert routing framework. StrLoRA first performs task-aware expert selection using the textual instruction to activate a sparse subset of relevant experts, reducing cross-task interference. It then applies token-wise expert weighting within this subset, where contribution weights are computed via cross-modal attention between local visual tokens and the global instruction representation. To maintain stability across the non-stationary stream, a routing-stability regularization aligns current routing distributions with a historical exponential moving average reference. Extensive experiments on a newly developed StrCVIT benchmark show that StrLoRA substantially outperforms existing methods, effectively enhancing model's abilities from continuously evolving data streams.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
Skim: Speculative Execution for Fast and Efficient Web Agents
arXiv:2605.16565v1 Announce Type: new
Abstract: Skim is a speculative execution framework for web agents that exploits the predictable structure of purpose-built websites. Today's web-agent expense is not intrinsic to the tasks but a property of how agents are composed: frontier-model inference, browser rendering, and ReAct-style planning are applied to every step of every task regardless of complexity. Skim's key observation is that websites enforce stable URL patterns, answer formats, and task-to-trajectory mappings across queries of the same type, so most queries can bypass these heavyweight components entirely. An offline profiler captures these patterns once per site. At runtime, Skim matches each query to a template, synthesizes the destination URL, and extracts the answer with a small model. A lightweight verifier gates each fast-path output against the query and schema; rare misspeculations cascade to the full agent, warm-started by the fast path's final URL to preserve upstream trajectory progress. Across standard web-agent benchmarks paired with three backboneagents (WebVoyager, AgentOccam, BrowserUse), Skim reduces median per-task cost by 1.9x and latency by 33.4% with no accuracy loss.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
Orth-Dion: Eliminating Geometric Mismatch in Distributed Low-Rank Spectral Optimization
arXiv:2605.16341v1 Announce Type: new
Abstract: Low-rank gradient compression reduces communication in distributed training by representing updates with rank-$r$ factors. Dion is a recent method that approximates Muon, a spectral optimizer that orthogonalizes momentum, using one step of power iteration followed by column normalization (rescaling each column of the right factor to unit length). This makes it compatible with fully sharded data parallel training, but it converges more slowly than full-rank spectral methods. We show that this gap is geometric: column normalization does not yield the rank-$r$ polar factor that Muon implicitly targets, so the resulting direction violates the dual-norm constraint of the low-rank spectral geometry, and the rate picks up an extra factor of $\sqrt{r}$ even though the low-rank approximation of the gradient itself is accurate. The same mismatch enters the smoothness term and the error-feedback recursion in the analysis, which has a knock-on effect on empirical performance. We propose Orth-Dion, which replaces column normalization with QR orthogonalization of the right factor. Under non-Euclidean smoothness, with $L_r$ the curvature constant along rank-$r$ directions, Orth-Dion attains rate $O(\sqrt{L_r/T})$, matching exact spectral methods at the same per-step communication cost as Dion. The proof removes the bounded-drift assumption common in prior error-feedback analyses via a self-consistent fixed-point argument, and uses a time-averaged contraction that only requires the error sequence to contract on average rather than at every step. Experiments on large-scale language model pre-training validate the predicted $\sqrt{r}$ scaling and show that Orth-Dion closes the convergence gap to Muon at Dion's communication cost.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
LoopQ: Quantization for Recursive Transformers
arXiv:2605.16343v1 Announce Type: new
Abstract: Looped language models (LoopLMs) improve parameter efficiency by recursively reusing Transformer blocks, enabling deeper computation under a fixed model size. However, this reuse makes LoopLMs more fragile under post-training quantization (PTQ). We present the first systematic study of quantization in LoopLMs and identify three challenges: distribution shift across roles, state reuse across loop transitions, and recursive error accumulation. To address these challenges, we propose LoopQ, a loop-aware PTQ framework that preserves a shared quantized backbone while introducing lightweight adaptations. LoopQ combines activation scaling, selective transformation, cross-loop state alignment, and trajectory-aware optimization to reduce distributional mismatch within loops and error accumulation across loops. Experiments across seven benchmarks show that, under W4A4 quantization, LoopQ improves average downstream accuracy by 68.8% and reduces average perplexity by 87.7% compared with the strongest static PTQ baseline.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
HPC-LLM: Practical Domain Adaptation and Retrieval-Augmented Generation for HPC Support
arXiv:2605.16347v1 Announce Type: new
Abstract: Modern scientific research increasingly depends on High-Performance Computing (HPC) infrastructures, yet many researchers face significant operational barriers when interacting with cluster environments, job schedulers, GPU resources, and parallel computing frameworks. General-purpose large language models (LLMs) provide useful coding assistance but often lack the domain-specific operational knowledge required for reliable HPC support. This paper presents HPC-LLM, a retrieval augmented and domain-adapted assistant designed to support common HPC workflows including Slurm scheduling, MPI execution, GPU utilization, filesystem management, and cluster troubleshooting. The proposed framework integrates automated documentation ingestion, dense retrieval, lightweight domain adaptation using QLoRA, and local inference within a modular orchestration pipeline. To support domain adaptation, we construct an HPC-oriented corpus from publicly available university HPC documentation, curated operational examples, and synthetic instruction-answer pairs generated from retrieved HPC content. The resulting dataset contains approximately 9,000 to 24,000 HPC-focused training examples spanning job scheduling, GPU computing, distributed training, storage systems, and cluster administration topics. We fine-tune Llama 3.1 8B using QLoRA and evaluate the resulting model against several open weight baselines under retrieval-augmented settings on JetStream2 infrastructure. Experimental results indicate that the adapted 8B model achieves performance comparable to substantially larger general-purpose models while operating under significantly lower GPU memory requirements and inference latency. In particular, the adapted model approaches the performance of Qwen 2.5 14B while requiring substantially fewer computational resources.
Fonte: arXiv cs.LG
Multimodal • Score 90
M$^2$FedAQI: Multimodal Federated Learning for Air Quality Prediction on Heterogeneous Edge Devices
arXiv:2605.16375v1 Announce Type: new
Abstract: Accurate air quality prediction is essential for public health, environmental monitoring, and industrial safety. However, most existing approaches rely on centralized learning paradigms, which introduce challenges related to scalability, privacy preservation, and communication overhead in distributed Internet of Things (IoT) environments. Moreover, current federated learning (FL) based solutions predominantly utilize unimodal data, limiting their capability to capture complex environmental patterns. To address these limitations, we propose M$^2$FedAQI, a lightweight multimodal federated framework for decentralized Air Quality Index (AQI) prediction across heterogeneous edge devices. The proposed framework integrates visual and tabular modalities through a feature modulation based fusion mechanism that enables efficient cross-modal interaction while maintaining low computational overhead. M$^2$FedAQI is evaluated on two benchmark datasets, PM25Vision and TRAQID, for both classification and regression tasks under centralized and federated settings. Experimental results demonstrate that M$^2$FedAQI consistently outperforms existing approaches, achieving improvements of up to 11.0\% in Accuracy, 3.53\% in AUC, 12.2\% in F1-score, and 18.0\% in $R^2$, while reducing MAE and RMSE by up to 25.4\% and 20.4\%, respectively, compared with the strongest baselines. Furthermore, deployment on heterogeneous edge devices demonstrates efficient resource utilization in terms of communication overhead, memory footprint, and computational cost. To enhance communication security, TLS-based authentication is incorporated to ensure secure client participation and protect the FL communication channel from unauthorized third-party access without modifying the underlying FL protocol.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
RAGA: Reading-And-Graph-building-Agent for Autonomous Knowledge Graph Construction and Retrieval-Augmented Generation
arXiv:2605.17072v1 Announce Type: new
Abstract: Existing LLM-driven knowledge graph (KG) construction methods predominantly employ stateless batch processing pipelines, exhibiting structural deficiencies in cross-chunk semantic relation capture, entity disambiguation, and construction process interpretability. These limitations undermine KG quality, retrieval precision, and deployment trust in high-stakes domains.
We propose RAGA (Reading And Graph-building Agent), an LLM-based autonomous KG construction and retrieval fusion framework. RAGA provides an atomic toolset supporting full KG lifecycle CRUD operations and embeds a Read-Search-Verify-Construct cognitive constraint into a ReAct tool loop. A KG-vector synchronization mechanism enables hybrid symbolic-vector retrieval, while evidence-anchored verification links every knowledge entry to its source text for auditable provenance.
Preliminary experiments on a subset of the QASPER scientific QA dataset indicate that RAGA's fusion retrieval outperforms zero-shot baselines, with KG integration providing measurable gains in both answer and evidence quality. The framework design and experimental baseline serve as a reference for agent-driven autonomous KG construction.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Concepts Worth Having: Refining VLM-Guided Concept Bottleneck Models with Minimal Annotations
arXiv:2605.16405v1 Announce Type: new
Abstract: Concept-bottleneck models (CBMs) are neural classifiers that compute predictions from high-level concepts extracted from the input. CBMs ensure stakeholders can understand the concepts -- and the predictions they entail -- by learning these from concept-level annotations, which are however seldom available. Recent CBM architectures work around this issue by obtaining annotations from Vision-Language Models (VLMs). While greatly broadening applicability, doing so can yield lower quality concepts and therefore less interpretable models. We strike for a middle ground by introducing Vision-plus-Human-guided CBM (VH-CBM), a hybrid approach that exploits both VLMs and a small amount of dense annotations. VH-CBM employs a Gaussian Process in the VLM's embedding space, which captures useful global information about the target domain, to propagate the expert's supervision to any target data point. Our empirical evaluation shows how VH-CBM predicts more accurate concepts than VLM-guided CBMs even when annotating as little as 1% of the data, while sporting better concept calibration and supporting active learning.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
GPU-Accelerated Deep Learning for Heatwave Prediction and Urban Heat Risk Assessment
arXiv:2605.16435v1 Announce Type: new
Abstract: Heatwaves are an important problem in cities, and climate change makes this problem more difficult. In this paper, we present a GPU-based deep learning framework for next-day prediction of urban thermal conditions and for heat risk assessment. The study was carried out in Sarajevo by using MODIS land surface temperature data and Open-Meteo forecast data. We tested several models, including convolutional models and spatiotemporal models. Among them, ConvLSTM with a mixed loss function gave the best results. The obtained values were MAE = 0.2293, RMSE = 0.3089, and R2 = 0.8877. The experiments also showed that results can be improved by using longer temporal series and additional meteorological variables. Since the framework was implemented on a GPU and trained with mixed precision, the execution time was reduced. Based on the predicted temperature fields, it was also possible to combine hazard information with exposure and vulnerability data in order to generate city heat risk maps. The proposed framework can be used as a practical basis for city heat analysis.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
ProxyKV: Cross-Model Proxy Pruning for Efficient Long-Context LLM Inference
arXiv:2605.16360v1 Announce Type: new
Abstract: Efficient long-context inference in Large Language Models (LLMs) is severely constrained by the Key-Value (KV) cache memory wall, yet existing pruning methods force a choice between
low-latency heuristics that sacrifice precision and high-precision reconstruction methods that incur prohibitive prefilling overhead. To bridge this scoring-cost--accuracy gap, we propose
ProxyKV, a cross-model proxy pruning framework that offloads importance scoring to a lightweight intra-family Small-Model Proxy executed asynchronously to the Large-Model Target. To bridge
the architectural gap between heterogeneous models, we design the HybridAxialMapper, which disentangles temporal feature extraction from cross-head alignment, together with a
Multi-Granularity Hybrid Loss that shifts the learning objective from rigid regression to relative ranking consistency. Across the Llama-3.1, Qwen-2.5, and Qwen-3 families spanning targets
from 7B up to 32B parameters on LongBench, SCBench, and RULER, ProxyKV matches KVZip on aggregate (recovering $\sim$$98.7\%$ of its mean accuracy) while delivering up to a $3.21\times$
prefilling speedup on Llama-3.1-8B (dual-GPU; $\sim$$1.5\times$ shared single-GPU) and sustaining the speedup at contexts up to 170k tokens on Qwen-2.5-7B.
Fonte: arXiv cs.LG
NLP/LLMs • Score 92
From Prompts to Protocols: An AI Agent for Laboratory Automation
arXiv:2605.16552v1 Announce Type: new
Abstract: Automating science laboratories enables faster, safer, more accurate, and more reproducible execution of protocols, accelerating the discovery and testing of new materials, drugs, and more. However, setting up and running autonomous labs requires coordinating numerous instruments and robots, forcing scientists to write code, manage configuration files, and navigate complex software infrastructure. We present an AI agent architecture that integrates large language models with laboratory orchestration, enabling scientists to interactively create and monitor automated lab protocols using natural language. Integrated into the Experiment Orchestration System (EOS), the AI agent operates under an agentic loop with automated validation and error correction, and supports the complete experimental lifecycle: creating protocols, running and monitoring both protocols and closed-loop optimization campaigns, and analyzing results. A visual graph editor renders protocols as interactive node-based diagrams synchronized with the AI agent's protocol representation, enabling seamless alternation between AI-assisted and manual protocol construction. Evaluated on three simulated automated labs spanning chemistry, biology, and materials science, the AI agent achieves a 97% first-attempt protocol generation success rate and an order of magnitude reduction in required interface actions.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
SignMuon: Communication-Efficient Distributed Muon Optimization
arXiv:2605.16311v1 Announce Type: new
Abstract: Distributed training of large neural networks is bottlenecked by full-precision gradient communication and by coordinatewise optimizers that ignore the matrix structure of weight tensors. We propose Sign-Muon, a 1-bit, matrix-aware optimizer that combines majority-vote sign aggregation from signSGD with the polar-step framework of Muon. Each worker forms a Muon-style direction by taking the polar factor of its momentum via a Newton--Schulz iteration, transmits only the entrywise signs, and aggregates by majority vote; an optional local polar step further enforces orthogonality at no extra communication cost.
Under spectral-norm smoothness and bounded-variance stochastic gradients, the spectral-norm normalized sign step yields an $\mathcal{O}(1/\sqrt{T})$ nonconvex rate for an $\ell_1$-based stationarity measure. With unimodal symmetric noise, majority vote across $M$ workers cuts the stochastic term by $1/\sqrt{M}$, matching signSGD. In the $\alpha$-$\beta$ model, distributed Sign-Muon needs only one integer sum-allreduce per iteration; all orthogonalization is local, giving a $32\times$ bandwidth reduction over float32 ($4\times$ for int8).
Across 330 CIFAR-10/ResNet-50 configurations Sign-Muon attains the best validation accuracy (92.15\%); its 4-GPU majority-vote variant reaches 92.02\% with 37\% less training time at matched effective batch. On nanoGPT, Sign-Muon achieves lower perplexity and better anytime performance than other sign-based baselines, with favorable weak-scaling up to 16 GPUs.
Fonte: arXiv cs.LG
Privacy/Security/Fairness • Score 85
PropGuard: Safeguarding LLM-MAS via Propagation-Aware Exploration and Remediation
arXiv:2605.16346v1 Announce Type: new
Abstract: LLM-based multi-agent systems (LLM-MAS) have become a promising paradigm for solving complex tasks through role specialization, tool use, memory, and collaborative reasoning. However, these interactions create new security risks that malicious instructions injected through messages, tools, or memories can propagate across agents and rounds, causing system-level compromise. Existing defenses largely rely on local filtering or graph-based anomaly detection, but they often fail to trace fine-grained propagation paths or remediate contaminated states without disrupting benign collaboration. We propose PropGuard, a propagation-aware framework for safeguarding LLM-MAS. PropGuard constructs a dual-view spatio-temporal graph that combines response-centric risk estimation with full-state evidence preservation. Guided by these risk priors, a GE-GRPO trained inspector sequentially explores the full-state graph to recover compact suspicious propagation subgraphs. PropGuard then verifies harmful propagation through subgraph-aware diagnosis and applies source-guided remediation to correct upstream contamination and replay affected downstream interactions. Experiments across four communication architectures and five attack settings demonstrate that PropGuard consistently lowers attack success while maintaining high task-level defense success, achieving a favorable effectiveness--efficiency trade-off.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
StatQAT: Statistical Quantizer Optimization for Deep Networks
arXiv:2605.17745v1 Announce Type: new
Abstract: Quantization is essential for reducing the computational cost and memory usage of deep neural networks, enabling efficient inference on low-precision hardware. Despite the growing adoption of uniform and floating-point quantization schemes, selecting optimal quantization parameters remains a key challenge, particularly for diverse data distributions encountered during training and inference. This work presents a novel statistical error analysis framework for uniform and floating-point quantization, providing theoretical insight into error behavior across quantization configurations. Building on this analysis, we propose iterative quantizers designed for arbitrary data distributions and analytic quantizers tailored for Gaussian-like weight distributions. These methods enable efficient, low-error quantization suitable for both activations and weights. We incorporate our quantizers into quantization-aware training and evaluate them across integer and floating-point formats. Experiments demonstrate improved accuracy and stability, highlighting the effectiveness of our approach for training low-precision neural networks.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Multi-Paradigm Agent Interaction in Practice:A Systematic Analysis of Generator-Evaluator, ReAct Loop,and Adversarial Evaluation in the buddyMe Framework
arXiv:2605.16821v1 Announce Type: new
Abstract: The rapid evolution of Large Language Model (LLM) agents has produced diverse interaction paradigms, yet few production systems integrate multiple paradigms within a unified architecture. This paper presents a systematic analysis of three principal agent interaction paradigms, including Multi-Agent Orchestration (Generator-Evaluator), ReAct Tool-Use Loops, and Memory-Augmented Interaction, as implemented in buddyMe, an open-source multi-model agent programming framework. We formalize a five-stage processing pipeline: Requirement Pre-Review -> Task Decomposition -> ReAct Execution -> Real-Execution Verification -> Adversarial Evaluation Discussion, and establish a six-dimensional evaluation schema with weighted scoring. Through four empirical case studies drawn from real-world deployment logs covering museum guide generation, scheduled weather tasks, and comprehensive tour planning, we draw three key conclusions. First, Generator-Evaluator pre-review detects requirement omissions in 20 percent of complex tasks, with 80 percent tasks passing initial inspection. Second, the ReAct loop ensures stable subtask execution but leads to around 30 percent redundant tool invocations. Third, adversarial Evaluator-Defender discussions reach consensus within 2-3 rounds for nearly 70 percent of scenarios, functioning mainly for content refinement rather than logical reversal. We additionally provide three Mermaid-based architectural diagrams and conduct cross-paradigm comparisons with CrewAI, AutoGen, LangGraph, MemGPT and A-Mem across six system dimensions. The research outcomes offer practical design guidelines for constructing stable and reliable multi-paradigm agent systems.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
GRID: Graph Representation of Intelligence Data for Security Text Knowledge Graph Construction
arXiv:2605.16714v1 Announce Type: new
Abstract: Security knowledge graphs can provide computable external memory for security agents, but constructing them from long-form cyber threat intelligence (CTI) remains difficult: LLMs often lack grounded security-domain knowledge, and end-to-end document-to-graph training is hard to supervise with cheap, stable rewards. We present GRID (Graph Representation of Intelligence Data), an end-to-end framework for security text knowledge graph construction. GRID first builds security-domain supervision from CTI articles by creating traceable article-graph alignments through graph extraction and knowledge-graph-conditioned text revision. It then turns document-to-graph learning into a scripted task bank combining four-option multi-select questions with triple-level regex matching targets, yielding more stable task-specific rewards than repeatedly scoring full graph outputs with an LLM judge. Using this supervision pipeline, we train two Qwen3-4B-Instruct-2507-based 4B extractors: a primary Task-bank Reward model and a secondary End2End Reward model with LLM-as-judge precision/recall rewards. On 249 CTI articles from GRID, CASIE, CTINexus, MalKG, and SecureNLP, the Task-bank Reward model with the ontology-guided GRID extraction pipeline reaches 84.62% source-averaged precision, 64.91% source-averaged recall, and 68.53% Avg F1, achieving the best source-averaged recall and near-top Avg F1 with lower token usage and deployment cost. The End2End Reward model reaches 76.91% precision, 53.85% recall, and 58.06% Avg F1. Further analyses show that task-bank rewards can be built once offline and reused across later post-training runs, outperforming online End2End LLM-as-judge reward and weaker alternatives such as Choice-only Reward and End2End SFT without RL.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Edge-AI-Driven Learning-to-Rank for Decentralized Task Allocation in Circular Smart Manufacturing
arXiv:2605.16433v1 Announce Type: new
Abstract: Task allocation in smart manufacturing systems needs to operate under decentralized decision-making, dynamic workloads, and shared resource constraints. In circular manufacturing settings, these challenges are further intensified by the need to balance operational efficiency with resource and energy sustainability. While learning-based approaches have been explored, many focus on predicting absolute performance metrics that do not necessarily translate into improved allocation outcomes, since decentralized assignment is governed by the relative ordering of candidate machines. This work proposes an Edge-AI-driven decentralized task allocation framework based on ranking-aware negotiation, where lightweight decision intelligence is embedded at the machine level to enable low-latency coordination without centralized control. The framework is developed progressively: a resource-aware heuristic first establishes the decentralized bidding structure, an Edge-AI-based regression model then provides learned local bid approximation, and a ranking-aware formulation finally reshapes the learning objective to align with the ordering-based nature of winner selection. Each machine evaluates incoming tasks using local information, including processing capability, queue state, and resource contention. The framework is evaluated via discrete-event simulation under high-load and tight-deadline scenarios using delay, deadline violations, throughput, and energy consumption. Results show improved delay and deadline adherence under high load, and enhanced energy efficiency under tighter constraints, leading to more resource-efficient operation aligned with circular manufacturing objectives. These findings demonstrate that aligning learning objectives with decentralized decision structures is critical for effective negotiation-driven task allocation.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Baba in Wonderland: Online Self-Supervised Dynamics Discovery for Executable World Models
arXiv:2605.16725v1 Announce Type: new
Abstract: Executable world models can be read, edited, executed, and reused for planning, but only if the program captures the environment's transition law rather than semantic shortcuts in its surface vocabulary. We study online executable world-model learning under prior misalignment, where an agent must induce state-dependent dynamics from interaction evidence alone, without rule descriptions, reward signals, or trustworthy lexical priors. We introduce Alice, a closed-loop system that treats failed candidate updates as structural signal: when a candidate explains a new transition but loses previously explained ones, the preservation conflict reveals dynamics that the current program had conflated. Alice refines these conflicts into hypothesis classes that both provide compact, class-stratified preservation counterexamples for update and guide frontier exploration toward transitions that are novel and underrepresented with respect to the current program. We evaluate Alice on Baba in Wonderland, a prior-misaligned variant of Baba Is You that preserves simulator dynamics while replacing semantically meaningful rule-property labels with unrelated words. Experiments show that Alice substantially improves executable world-model learning under prior misalignment, and ablations show that both class refinement and class-aware exploration contribute.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Sustainable Intelligence for the Wild: Democratizing Ecological Monitoring via Knowledge-Adaptive Edge Expert Agents
arXiv:2605.16671v1 Announce Type: new
Abstract: Rapid biodiversity loss underscore the urgency of effective monitoring, yet manual surveys remain resource-intensive. While on-device AI offers a scalable alternative, its performance in the wild is often challenged by environmental variability. Current methods rely heavily on cloud resource, which requires continuous uploading of field data for model retraining. This approach is unsuitable for remote deployments because it consumes limited power and network connectivity. To address these constraints, this research proposes a shift from model adaptation to knowledge adaptation. We introduce an architecture that separates visual perception from reasoning, combining a visual encoder with a dynamic knowledge base. We uses an explicit knowledge base to replace implicitly encoding expert knowledge into model parameters. This method also supports knowledge sustainability by preserving expert insights in a structured form. Through cross-disciplinary collaboration with biologists and Indigenous communities, this work advances ethical AI co-development, fostering responsible and culturally informed ecosystem management.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Federated Nested Learning: Collaborative Training of Self-Referential Memories for Test-Time Adaptation
arXiv:2605.16350v1 Announce Type: new
Abstract: We rethink Federated Learning (FL) from a nested learning perspective, framing the core challenge as how to collaboratively learn optimization rules, not just static models, to tackle Non-IID client data. To address this, we propose Federated Nested Learning (FedNL), a novel framework that reformulates FL as a three-level nested optimization system. FedNL embeds Titans-based linear attention into FL, enabling clients to perform lightweight, zero-shot test-time adaptation by treating a delta rule as an online gradient step. Experiments on Non-IID MMLU and long-context benchmarks show that FedNL achieves competitive performance in short-context reasoning, enhances the performance of long-context retrieval and streaming Cross-Entropy, and maintains constant inference memory.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Physics-Guided Geometric Diffusion for Macro Placement Generation
arXiv:2605.16451v1 Announce Type: new
Abstract: Macro placement is a pivotal stage in VLSI physical design, fundamentally determining the overall chip performance. Recent data-driven placement methods have demonstrated significant potential, yet they often struggle to handle sequential dependencies and to balance topological connectivity with physical constraints. To bridge this gap, we propose MacroDiff+, a physics-guided geometric diffusion framework. Specifically, we design a dual-domain denoising architecture that couples topological connectivity encoded by heterogeneous GNNs with global geometric context modeled by a Transformer. Furthermore, we introduce Physics-Guided Sampling, an inference strategy that actively steers the generation using explicit gradients to ensure both statistical plausibility and physical validity. On the ISPD2005 MMS benchmarks, MacroDiff+ outperforms state-of-the-art baselines with a 6.1-6.2% reduction in wirelength. Notably, it exhibits superior stability and scalability on large-scale designs where prior methods fail to converge. The source code is available at https://github.com/jhy00n/MacroDiff-plus.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
E-PMQ: Expert-Guided Post-Merge Quantization with Merged-Weight Anchoring
arXiv:2605.16882v1 Announce Type: new
Abstract: Low-resource deployment constraints have made model quantization essential for deploying neural networks while preserving performance. Meanwhile, model merging has become an increasingly practical low-resource strategy for integrating multiple task- or domain-specialized experts into a single model without joint training or multi-model serving. Together, quantization and model merging enable an efficient low-resource deployment pipeline by integrating multiple experts into one low-bit model. We formulate this setting as Post-Merge Quantization (PMQ). We show that directly applying post-training quantization (PTQ) to a merged model is unreliable because two distinct deviations are coupled: the quantization deviation introduced by low-bit reconstruction and the expert-relative merging deviation inherited from model merging. To mitigate these deviations, we propose E-PMQ, an expert-guided PMQ framework that uses source expert weights to provide expert- guided output targets during layer-wise calibration, together with merged-weight anchoring to stabilize the calibration and preserve the integrated behavior of the merged model. On CLIP-ViT-B/32 eight-task merging, E-PMQ improves 4-bit GPTQ from 65.0% to 73.6% under Task Arithmetic and from 69.1% to 74.8% under TIES-Merging. On harder settings, E-PMQ improves GPTQ from 34.8% to 76.7% on 20-task CLIP-ViT-L/14 and from 78.26% to 83.34% on FLAN-T5- base GLUE. These results demonstrate that E-PMQ enables effective post-merge quantization and low-bit deployment.
Fonte: arXiv cs.CL
RL • Score 85
Reducing Credit Assignment Variance via Counterfactual Reasoning Paths
arXiv:2605.16302v1 Announce Type: new
Abstract: Reinforcement learning for multi-step reasoning with large language models (LLMs) often relies on sparse terminal rewards, leading to poor credit assignment conditions where the final feedback is evenly propagated across all intermediate decisions. This results in high gradient variance, unstable training, and numerous ineffective updates, ultimately causing the model to fail and preventing sustained improvement. We introduce a counterfactual comparison-based credit assignment framework, which samples multiple reasoning trajectories under the same input. By treating their differences as an implicit approximation of alternative decisions, we construct an implicit process-level advantage estimator that transforms sparse terminal rewards into step-sensitive learning signals. Based on this, we propose Implicit Behavior Policy Optimization (IBPO), which significantly improves training stability and performance upper bounds on mathematical and code reasoning benchmarks, pointing to a promising direction for unlocking the performance potential of LLMs.
Fonte: arXiv cs.LG
NLP/LLMs • Score 90
HyDRA: Hybrid Dynamic Routing Architecture for Heterogeneous LLM Pools
arXiv:2605.17106v1 Announce Type: new
Abstract: Production LLM deployments increasingly maintain heterogeneous model pools spanning order-of-magnitude cost differences. Existing routers make binary strong-vs-weak decisions and couple learned parameters to specific model identities, requiring retraining whenever the catalog changes. We present HyDRA (Hybrid Dynamic Routing Architecture), a framework that predicts fine-grained, multi-dimensional capability requirements per query and matches them against configuration-defined model profiles via shortfall matching. A ModernBERT encoder with K=4 independent sigmoid heads scores each query along reasoning, code generation, debugging, and tool use; a shortfall-matching algorithm then selects the cheapest model whose capabilities meet the predicted requirements. The deployed predictor runs at 86 ms median CPU inference latency in production, and is fully decoupled from the model catalog -- adding or removing models requires only a configuration change, with zero retraining. On SWE-Bench Verified (5-model pool: GPT-5.4-mini, Claude Haiku 4.5, GPT-5.3 Codex, Claude Sonnet 4.6, GPT-5.4), HyDRA's tunable shortfall threshold spans three regimes: peak-quality exceeds the always-strong Claude Sonnet 4.6 baseline (75.4% vs. 74.2% resolution) at 12.9% cost savings; iso-quality matches Sonnet at 54.1% cost savings, a 6x improvement over our prior in-house binary router at 9.1%; aggressive pushes savings to 72.5% for a 3.2-point quality trade. Results generalize across LiveCodeBench, BigCodeBench, and tau-bench. HyDRA is deployed to all users in GitHub Copilot's VS Code Chat auto-mode and -- to our knowledge for the first time in the LLM routing literature -- demonstrates language-invariant routing across CJK, European, and other script families.
Fonte: arXiv cs.CL
MLOps/Systems • Score 85
PRISMat: Policy-Driven, Permutation-Invariant Autoregressive Material Generation
arXiv:2605.16612v1 Announce Type: new
Abstract: Rapid identification of candidate materials with target properties has become a key task in materials science. Machine learning has emerged as an alternative to physics-based simulation, offering a faster and cheaper way to filter materials based on their stability and other target properties, reducing the number of candidates that reach the costly synthesis stage. Recently, Large Language Models (LLMs) have been applied to this role, but these models are parameter-heavy and computationally expensive both during training and at inference time, making them unsuitable for high-throughput tasks. This inefficiency stems from both the large over-parameterization of language models and the difficulty of framing material generation as a sequence learning problem. In this paper, we present PRISMat, a cost-effective, permutation-invariant model, which addresses these limitations. We show that PRISMat, despite taking less time for inference, is able to outperform LLMs in generating crystal slabs conditioned on critical materials' surface properties. In targeted material discovery, we achieve mean absolute errors of 0.188 eV/A$^2$ and 2.79 eV for cleavage energy and work function tasks, respectively, reducing the error of the next best model by 4$\times$.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
A Theory of Training Profit-Optimal LLMs
arXiv:2605.16430v1 Announce Type: new
Abstract: Scaling LLMs requires tremendous computational resources, and recent advances in AI have gone hand in hand with massive amounts of capital expenditure. While it is established that scaling up LLMs reliably increases model quality (quantified in terms of loss or downstream evaluations), it is unclear how these quality improvements translate to potential revenue, and whether revenue increases would offset costs of larger-scale training and inference. In this work, we develop an economic model for characterizing the rational behavior of an LLM training firm by combining scaling laws with microeconomic theory. Under our model of firm behavior, LLM quality can be increased with more parameters and training tokens, leading to more potential adoption by consumers, who each have a quality threshold for using the LLM. On the other hand, additional parameters and training tokens both incur additional costs. We analyze the profit maximization problem for this model under compute-bound and data-bound regimes. In the compute-bound regime, optimal model size and token budget track hardware efficiency $E$ (FLOPs/\$) at a near-linear rate; total training cost then scales sub-quadratically in $E$. Data efficiency improvements incentivize larger models and training expenditure. When we are limited to $D$ data, profit-optimal training expenditure scales as $D^2/E$, i.e, increase with data and decreases with hardware efficiency (as well as data efficiency). Finally, we analyze practical trends in training expenditure: current trends are consistent with our most permissive model variants in the compute-bound regime, but are not profit-optimal in the data-bound regime or assuming hardware advances will stall. Overall, our results provide a theory of profit-optimal LLM training, providing a foundation for engaging critically with industry statements and supporting long-term economic decision making.
Fonte: arXiv cs.LG
Privacy/Security/Fairness • Score 85
AgentWall: A Runtime Safety Layer for Local AI Agents
arXiv:2605.16265v1 Announce Type: new
Abstract: The safety of autonomous AI agents is increasingly recognized as a critical open problem. As agents transition from passive text generators to active actors capable of executing shell commands, modifying files, calling APIs, and browsing the web, the consequences of unsafe or adversarially manipulated behavior become immediate and tangible. Existing AI safety work has focused primarily on model alignment and input filtering, but these approaches do not address what happens at the moment an agent's intent becomes a real action on a real machine. This gap is especially acute in local environments, where developers run agents against their own filesystems, credentials, and infrastructure with little runtime control. This paper introduces AgentWall, a runtime safety and observability layer for local AI agents. AgentWall intercepts every proposed agent action before it reaches the host environment, evaluates it against an explicit declarative policy, requires human approval for sensitive operations, and records a complete execution trail for audit and replay. It is implemented as a policy-enforcing MCP proxy and native OpenClaw plugin, working across Claude Desktop, Cursor, Windsurf, Claude Code, and OpenClaw with a single install command. We present the design, architecture, threat model, and policy model of AgentWall, and demonstrate 92.9% policy enforcement accuracy with sub-millisecond overhead across 14 benchmark tests. AgentWall is open-source at https://github.com/agentwall/Agentwall.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Nested Spatio-Temporal Time Series Forecasting
arXiv:2605.16447v1 Announce Type: new
Abstract: Spatiotemporal forecasting is critical for real-world applications like traffic management, yet capturing reliable interactions remains challenging under noisy and non-stationary conditions. Existing methods primarily rely on historical spatial priors, often failing to account for evolving temporal correlations and suffering from systematic errors. In this work, we propose a nested forecasting framework that couples future macro-level regional trends with micro-level historical observations, enabling top-down guidance from abstract future representations for fine-grained forecasting. Specifically, we employ a spectral clustering-based approach to construct semantically coherent regions, providing both theoretical and empirical evidence that this representation effectively filters systematic noise while preserving essential trends. Building on this, we develop a progressive coarse-to-fine predictor to integrate these representative features into the inference process. This enables the model to leverage trend predictions to anticipate dynamic anomalies, such as periodic offsets, in advance. Furthermore, extensive experiments on multiple high-dimensional datasets demonstrate that our method consistently outperforms state-of-the-art baselines, validating the effectiveness of future macro-guided nested forecasting.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
A note on connections between the F\"ollmer process and the denoising diffusion probabilistic model
arXiv:2605.18040v1 Announce Type: new
Abstract: The F\"ollmer process is a Brownian motion conditioned to have a pre-specified distribution at time 1. This process can be interpreted as an "augmented" time-compressed version of the reverse stochastic differential equation (SDE) for the denoising diffusion probabilistic model (DDPM). While this fact has been indirectly used to analyze DDPM sampling errors via discretization of the reverse SDE, connections between direct discretization of the F\"ollmer process and the DDPM sampler have not yet been fully explored. This note aims to clarify this point while surveying relevant results from existing work. We show that discretized F\"ollmer processes give natural hyper-parameter settings of the DDPM sampler. Moreover, this allows us to systematically recover state-of-the-art results on DDPM sampling error bounds with slight improvements.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
MiniGPT: Rebuilding GPT from First Principles
arXiv:2605.17398v1 Announce Type: new
Abstract: This paper presents MiniGPT, a compact from-scratch implementation of GPT-style autoregressive language modeling in PyTorch. The aim is to rebuild the core GPT pipeline from first principles after studying the design of nanoGPT by Andrej Karpathy, while keeping the model and training code independently written in a single notebook. MiniGPT implements token and positional embeddings, causal multi-head self-attention, pre-LayerNorm Transformer blocks, residual connections, feed-forward MLP layers, next-token cross-entropy training (teacher forcing), validation tracking, checkpoint selection, and autoregressive text generation. This paper evaluates the implementation on Tiny Shakespeare dataset using character-level tokenization. A baseline 0.83M-parameter model reaches a validation loss of 1.7236 after 3000 training iterations. A stronger 10.77M-parameter configuration, using a larger context length and improved training settings, reaches a best validation loss of 1.4780 and generates text with recognizable Shakespeare-style dialogue structure. MiniGPT does not introduce a new language-model architecture. Instead, it documents a clear and reproducible implementation path from raw text to trained character-level generation, including design choices, training behavior, generation quality, and practical limitations.
Fonte: arXiv cs.CL
Multimodal • Score 85
How Many Visual Tokens Do Multimodal Language Models Need? Scaling Visual Token Pruning with F^3A
arXiv:2605.16359v1 Announce Type: new
Abstract: Vision-language models improve perception by feeding increasingly long visual token sequences into language backbones, but the resulting inference cost raises a basic scaling question: as multimodal models grow, how many visual tokens are actually needed, and how should they be allocated under a fixed visual token budget? Existing training-free pruning methods typically answer this with one-shot proxies such as decoder attention, visual similarity, or conditional diversity. We argue that visual token pruning is better viewed as task-conditioned evidence search, especially under aggressive compression and across model scales. We propose F^3A, a training-free router for visual token pruning that operates before the language model consumes image tokens. F^3A builds lightweight question-conditioned cues, matches them to visual-grid tokens through frozen sparse sensing heads, and allocates a fixed vision token budget via coarse evidence localization, local refinement, coverage-preserving competition, and recovery of under-covered regions. It requires no model training, no extra LLM forward pass and preserves the original multimodal prompting and decoding pipeline.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
Statistical Limits and Efficient Algorithms for Differentially Private Federated Learning
arXiv:2605.18656v1 Announce Type: new
Abstract: Federated Learning is a leading framework for training ML and AI models collaboratively across numerous user devices or databases. We study the trade-offs among estimation accuracy, privacy constraints, and communication cost for differentially private (DP) federated M estimation. The two standard methods in the literature are FedAvg, which may suffer from high federation bias, and FedSGD, which can incur high communication cost. Aimed at improving accuracy at a reduced communication cost, we propose FedHybrid, which uses FedSGD starting with an improved initialization by the FedAvg estimator. We propose FedNewton, which averages local Newton iterations to reduce bias in FedAvg, achieving an estimation accuracy comparable to FedSGD with much fewer communication rounds when the number of clients grows sufficiently slowly. We establish finite sample upper bounds on the mean-squared error rates of the DP versions of these estimators as functions of the number of clients, local sample sizes, privacy budget, and number of iterations. We further derive a minimax lower bound on the MSE of any iterative private federated procedure that provides a benchmark to assess the optimality gap of these methods. We numerically evaluate our methods for training a logistic regression and a neural network on the computer vision datasets MNIST and CIFAR-10.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
NGM: A Plug-and-Play Training-Free Memory Module for LLMs
arXiv:2605.16893v1 Announce Type: new
Abstract: Recent studies introduce conditional memory modules that decouple knowledge storage from neural computation, enabling more direct knowledge access. Compared to MoE, which relies on dynamic computation paths, explicit lookup provides a more efficient knowledge retrieval mechanism. However, these approaches still depend on learned memory embeddings, requiring additional training and limiting flexibility. To address this, we propose N-gram Memory (NGM), a training-free, plug-and-play module composed of a Causal N-Gram Encoder and a Cosine-Gated Memory Injector. The Causal N-Gram Encoder directly averages the pretrained token embeddings of the backbone model to construct N-gram representations, thereby eliminating the need to train separate N-gram embeddings from scratch. This design requires neither an additional memory table nor a retrieval pipeline. The Cosine-Gated Memory Injector then uses a non-parametric cosine gate with ReLU to modulate the retrieved embeddings into the contextual representations. We evaluate NGM on the Qwen3 series from 0.6B to 14B across eight benchmarks. NGM improves average performance by 0.5 to 1.2 points, with particularly clear gains on code generation and knowledge-intensive tasks (e.g., +3.0 on LiveCodeBench and +3.03 on GPQA for Qwen3-14B). Moreover, NGM also improves performance in multimodal benchmarks (e.g., MMStar +1.53 on Qwen3-VL-2B).
Fonte: arXiv cs.AI
NLP/LLMs • Score 90
MADP: A Multi-Agent Pipeline for Sustainable Document Processing with Human-in-the-Loop
arXiv:2605.17159v1 Announce Type: new
Abstract: Document processing automation remains a critical challenge in enterprise environments, where traditional manual approaches are labor-intensive and error-prone. We present MADP, a multi-agent architecture that addresses the challenge of automating document processing in enterprise settings by combining deep learning-based classification and parsing with large language model extraction, while maintaining accuracy through selective human validation. Our system integrates five specialized agents--Classificator, Splitter, Parser, Extraction, and Validator--with a Human-in-the-Loop (HITL) mechanism and a novel Prompt Fine Tuning with Feedback Inheritance (PFTFI) approach. The operational analysis on a production use-case scenario of 100,000 invoices per year indicates a potential reduction of Full-Time Equivalent (FTE) requirements by approximately 70%. Production deployment on 955 real-world documents processed through January 2026 achieves a 97.0% full-pipeline automation rate, with only 3% requiring non-AI fallback. Ablation evaluation on a stratified 100-document subset (5 documents per each of 20 supplier/document-type categories) demonstrates that the full MADP configuration with Human-in-the-Loop supervision attains 98.5% document-level accuracy. Additionally, we present a comprehensive sustainability analysis showing that our hybrid AI+HITL approach reduces CO2 emissions by 69%, energy consumption by 69%, and water usage by 63% compared to traditional manual processing. Benchmark comparisons of multiple LLM backends (Granite-Docling, Mistral-Small, DeepSeek-OCR) provide practical insights for deployment in production environments.
Fonte: arXiv cs.AI
MLOps/Systems • Score 75
Scaling Accessible Mathematics on arXiv: HTML Conversion and MathML 4
arXiv:2605.16562v1 Announce Type: new
Abstract: We report on the ongoing development of arXiv's HTML Papers offering, available on every new TeX/LaTeX submission since its initial release in 2023.
The main highlights from 2025 and early 2026 are:
(i) community-driven improvements to HTML fidelity and service health, with roughly half of 6,000 user reports resolved;
(ii) corpus-scale conversion work aimed at 90% error-free HTML (currently 75%);
(iii) initial MathML 4 Intent annotations for accessible speech output;
(iv) an in-progress Rust port of LaTeXML, reducing compute costs and enabling faster previews on submission.
The arXiv HTML Papers project remains experimental, but is gradually maturing as we better understand the needs of arXiv's readers and the technical opportunities presented by new standards and by advances in programming languages and AI.
Fonte: arXiv cs.CL
MLOps/Systems • Score 85
Systematic Optimization of Real-Time Diffusion Model Inference on Apple M3 Ultra
arXiv:2605.16259v1 Announce Type: new
Abstract: While real-time image generation using diffusion models has advanced rapidly on NVIDIA GPUs, systematic optimization research on non-CUDA platforms such as Apple Silicon remains extremely limited. In this study, we conducted comprehensive optimization experiments across 10 phases targeting the Apple M3 Ultra (60-core GPU, 512 GB unified memory) with the goal of achieving real-time camera img2img transformation. We explored a wide range of techniques including CoreML conversion, quantization, Token Merging, Neural Engine utilization, compact model exploration, frame interpolation, kNN search-based synthesis, pix2pix-turbo, optical flow frame skipping, and knowledge distillation, quantitatively evaluating the effectiveness of each approach. Ultimately, by combining CoreML conversion of the distillation-specialized model SDXS-512 with a 3-thread camera pipeline, we achieved real-time camera img2img transformation at 22.7 FPS at 512x512 resolution. The primary contribution of this work is the systematic demonstration that optimization insights established for CUDA are not necessarily effective on Apple Silicon's unified memory architecture. We reveal an optimization landscape fundamentally different from that of NVIDIA GPUs -- including the absence of speedup from quantization, the ineffectiveness of parallel inference, and the unsuitability of the Neural Engine for large-scale models -- and provide practical guidelines for diffusion model inference on Apple Silicon.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Reliability and Effectiveness of Autonomous AI Agents in Supply Chain Management
arXiv:2605.17036v1 Announce Type: new
Abstract: This paper studies autonomous generative AI agents in multi-echelon supply chains using the MIT Beer Game. We identify four inference-time levers that shape performance: model selection, policies and guardrails, centralized data sharing, and prompt engineering. Model capability is the dominant factor: an out-of-the-box reasoning model exceeds human-level performance, and optimized reasoning models reduce costs by up to 67% relative to human teams. However, strong average performance masks substantial reliability risks. We introduce the agent bullwhip effect, the amplification of decision unreliability across echelons, manifesting along two dimensions: decision variance increases both across facilities at the same point in time and within the same facility across time. We develop a mathematical framework showing that this phenomenon is inherent to multi-agent systems that involve coordination and information delays, and we demonstrate that repeated sampling fails to meaningfully reduce it. To address this limitation, we propose a Group Relative Policy Optimization (GRPO)-based reinforcement-learning post-training framework that trains a shared base LLM using system-level supply-chain rewards. GRPO post-training substantially reduces tail events, curtails agent bullwhip, and improves the reliability of autonomous supply-chain agents.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Response-Conditioned Parallel-to-Sequential Orchestration for Multi-Agent Systems
arXiv:2605.15573v1 Announce Type: new
Abstract: Multi-agent systems can solve complex tasks through collaboration between multiple Large Language Model agents. Existing collaboration frameworks typically operate in either a parallel or a sequential mode. In the parallel mode, agents respond independently to queries followed by aggregation of responses. In contrast, sequential systems allow agents to communicate via a directed topology and refine one another step by step. However, both modes are inadequate for achieving the desired objectives of minimizing communication and latency while simultaneously maximizing the accuracy of the final response. In this work, we introduce a hybrid paradigm called Nexa, a trainable response-conditioned policy that bridges the gap between the two modes. Nexa begins with a parallel execution stage, embeds the resulting responses into a shared semantic space, and then predicts a sparse directed acyclic communication graph. If the graph is empty, the system remains purely parallel; if it is non-empty, the system performs one sequential message propagation. The policy is a lightweight transformer model, and the method avoids the need for external LLM judges or reward models, as well as hand-crafted test-time topology search. We formalize this hybrid execution problem, show that the resulting graph is acyclic by construction, and that the framework strictly subsumes pure parallel execution, and present a training procedure based on policy-gradient optimization. Results demonstrate that the response-conditioned policy learned by Nexa under one setting can be reused when the number of agents, the task, or the underlying agent changes, thus emphasizing the generalizability of the learned communication policy.
Fonte: arXiv cs.CL
Privacy/Security/Fairness • Score 85
Verifiable Agentic Infrastructure: Proof-Derived Authorization for Sovereign AI Systems
arXiv:2605.15228v1 Announce Type: new
Abstract: Modern cloud and enterprise systems rely on identity-centric authorization, assuming that callers possessing valid credentials are safe to execute commands. The emergence of autonomous AI agents invalidates this assumption: agents can generate syntactically valid but semantically unsafe actions, making standing privileges a significant operational risk. This risk becomes especially acute in sovereign AI systems, where autonomous agents may interact with cloud infrastructure, regulated data, financial workflows, and national-scale digital services. Governed mutation substrates reduce this risk by interposing on agent actions: agents submit intents, infrastructure evaluates context and policy, and execution is mediated. However, this shifts the trust boundary: how can the decision to authorize an intent be made verifiable, distributed, and replayable?
We introduce a Distributed Trust Framework (DTF), a verification framework for governed mutation systems that computes execution authority from structured, verifiable artifacts. DTF introduces a Justification Proof to encode the admissibility basis of an action, a consensus model for independent evaluation, an ephemeral Execution Identity derived from the approved proof, and an append-only Evidence Chain that preserves the authorization lifecycle. Under stated substrate assumptions, this architecture enforces a compact authorization invariant: no high-stakes execution without a proof object, no derived authority without consensus, and no valid mutation detached from evidence.
We define the model, instantiate it over an OpenKedge-based governed mutation substrate, and show how it maps onto cloud-native environments. By shifting authorization from standing identity to proof-derived authority, DTF provides an infrastructure foundation for making agentic execution governable, auditable, and bounded in sovereign AI deployments.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
DiLA: Disentangled Latent Action World Models
arXiv:2605.15725v1 Announce Type: new
Abstract: Latent Action Models (LAMs) enable the learning of world models from unlabeled video by inferring abstract actions between consecutive frames. However, LAMs face a fundamental trade-off between action abstraction and generation fidelity. Existing methods typically circumvent this issue by using two-stage training with pre-trained world models or by limiting predictions to optical flow. In this paper, we introduce DiLA, a novel Disentangled Latent Action world model that aims to resolve this trade-off via content-structure disentanglement. Our key insight is that disentanglement and latent action learning are co-evolving: the predictive bottleneck inherent in latent action learning serves as a driving force for disentanglement, compelling the model to distill spatial layouts into the structure pathway while offloading visual details to a separate content pathway for generation. This synergy yields a continuous, semantically structured latent action space without compromising generative quality. DiLA achieves superior results in video generation quality, action transfer, visual planning, and manifold interpretability. These findings establish DiLA as a unified framework that simultaneously achieves high-level action abstraction and high-fidelity generation, advancing the frontier of self-supervised world model learning.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Formal Methods Meet LLMs: Auditing, Monitoring, and Intervention for Compliance of Advanced AI Systems
arXiv:2605.16198v1 Announce Type: new
Abstract: We examine one particular dimension of AI governance: how to monitor and audit AI-enabled products and services throughout the AI development lifecycle, from pre-deployment testing to post-deployment auditing. Combining principles from formal methods with SoTA machine learning, we propose techniques that enable AI-enabled product and service developers, as well as third party AI developers and evaluators, to perform offline auditing and online (runtime) monitoring of product-specific (temporally extended) behavioral constraints such as safety constraints, norms, rules and regulations with respect to black-box advanced AI systems, notably LLMs. We further provide practical techniques for predictive monitoring, such as sampling-based methods, and we introduce intervening monitors that act at runtime to preempt and potentially mitigate predicted violations. Experimental results show that by exploiting the formal syntax and semantics of Linear Temporal Logic (LTL), our proposed auditing and monitoring techniques are superior to LLM baseline methods in detecting violations of temporally extended behavioral constraints; with our approach, even small-model labelers match or exceed frontier LLM judges. Our predictive and intervening monitors significantly reduce the violation rates of LLM-based agents while largely preserving task performance. We further show through controlled experiments that LLMs' temporal reasoning shows a pronounced degradation in accuracy with increasing event distance, number of constraints, and number of propositions.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
RTL-BenchMT: Dynamic Maintenance of RTL Generation Benchmark Through Agent-Assisted Analysis and Revision
arXiv:2605.15537v1 Announce Type: new
Abstract: This paper introduces RTL-BenchMT, an agentic framework for dynamically maintaining RTL generation benchmarks. Large Language Models (LLMs) assisted automated RTL generation is one of the most important directions in EDA research. However, current RTL benchmarks face two critical challenges: (1) flawed cases in the benchmarks and (2) overfitting to the benchmarks. Both challenges are difficult to resolve purely by manual engineering effort. To address these issues and systematically reduce human maintenance costs, we propose an automated agentic framework, RTL-BenchMT. RTL-BenchMT focuses on two key applications: (1) automatically identifying and revising flawed benchmark cases and (2) automatically detecting and updating overfitting cases. With the assistance of RTL-BenchMT, we conduct a thorough, in-depth analysis of flawed and overfitting cases and produce a refined benchmark suite that will be open-sourced to the community.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
DeepSlide: From Artifacts to Presentation Delivery
arXiv:2605.15202v1 Announce Type: new
Abstract: Presentations are a primary medium for scholarly communication, yet most AI slide generators optimize the artifact (a visually plausible deck) while under-optimizing the delivery process (pacing, narrative, and presentation preparation). We present DeepSlide, a human-in-the-loop multi-agent system that supports preparing the full presentation process, from requirement elicitation and time-budgeted narrative planning, to evidence-grounded slide--script generation, attention augmentation, and rehearsal support. DeepSlide integrates (i) a controllable logical-chain planner with per-node time budgets, (ii) a lightweight content-tree retriever for grounding, (iii) Markov-style sequential rendering with style inheritance, and (iv) sandboxed execution with minimal repair to ensure renderability. We further introduce a dual-scoreboard benchmark that cleanly separates static artifact quality from dynamic delivery excellence. Across 20 domains and diverse audience profiles, DeepSlide matches strong baselines on artifact quality while consistently achieving larger gains on delivery metrics, improving narrative flow, pacing precision, and slide--script synergy with clearer attention guidance.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Logical Grammar Induction via Graph Kolmogorov Complexity: A Neuro-Symbolic Framework for Self-Healing Clinical Data Integrity
arXiv:2605.15242v1 Announce Type: new
Abstract: The reliability of Healthcare Information Systems (HIS) is frequently compromised by human-induced data entry errors, which existing statistical anomaly detection methods fail to distinguish from legitimate clinical extremes. This paper proposes Logic-GNN, a novel neuro-symbolic framework that treats clinical records as a structured ``private language'' governed by latent logical games. By integrating Temporal Graph Neural Networks (TGNN) with Graph Kolmogorov Complexity, we induce a symbolic grammar that represents the underlying logic of medical interactions. We define anomalies as ``grammatical violations'' that cause a significant expansion in the Minimum Description Length (MDL) of the clinical graph. Evaluated on the Sina System dataset (2M+ records), Logic-GNN achieves an F1-score of 0.94, outperforming state-of-the-art baselines by 12\% in distinguishing between life-threatening medical outliers and data corruption. Our approach introduces a self-healing mechanism that suggests logical corrections to maintain data integrity in real-time HIS environments.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
RecMem: Recurrence-based Memory Consolidation for Efficient and Effective Long-Running LLM Agents
arXiv:2605.16045v1 Announce Type: new
Abstract: Memory systems often organize user-agent interactions as retrievable external memory and are crucial for long-running agents by overcoming the limited context windows of LLMs. However, existing memory systems invoke LLMs to process every incoming interaction for memory extraction, and such an eager memory consolidation scheme leads to substantial token consumption. To tackle this problem, we propose RecMem by rethinking when memory consolidation should be conducted. RecMem stores incoming interactions in a subconscious memory layer and encode them using lightweight embedding models for retrieval. LLMs are only invoked to extract episodic and semantic memory when sustained recurrence are observed for semantically similar interactions. Such recurrence-based consolidation works because these interactions correspond to a semantic cluster with rich information and thus are worth extraction and summarization. To improve accuracy, RecMem also incorporates a semantic refinement mechanism that recovers the fine-grained facts omitted by memory extraction. Experiments show that RecMem reduces the memory construction token cost of three SOTA memory systems by up to 87% while exceeding their accuracy.
Fonte: arXiv cs.CL
Vision • Score 92
ElasticDiT: Efficient Diffusion Transformers via Elastic Architecture and Sparse Attention for High-Resolution Image Generation on Mobile Devices
arXiv:2605.15684v1 Announce Type: new
Abstract: The Diffusion Transformer (DiT) architecture is the state-of-the-art paradigm for high-fidelity image generation, underpinning models like Stable Diffusion-3 and FLUX.1. However, deploying these models on resource-constrained mobile devices entails prohibitive computational and memory overhead. While efficiency-driven approaches like Linear-DiT and static pruning alleviate bottlenecks, they often incur quality degradation. Unlike cloud environments, mobile constraints require a single-model paradigm that dynamically balances fidelity and latency. We introduce ElasticDiT, which achieves this dynamic trade-off by adjusting spatial compression ratios and DiT block depths. By integrating Shift Sparse Block Attention (SSBA) and a Tiny DWT-Distilled VAE (T-DVAE), ElasticDiT reduces inference latency and memory footprint while maintaining image quality. Experiments confirm that ElasticDiT effectively covers a wide range of fidelity-latency trade-offs within a single set of parameters. By jointly adjusting compression and depth, a single ElasticDiT model can be reconfigured on-the-fly to outperform task-specific baselines. Specifically, our flex lite variant achieves an HPS of 32.87, surpassing the Flux model, while maintaining competitive quality at 84.16 percent average sparsity through SSBA. Furthermore, the plug-and-play T-DVAE provides SD3-level reconstruction with only 1/8x the computational cost of standard VAEs, and Flow-GRPO boosts semantic alignment (GenEval: 66.93 to 73.62). These results demonstrate that ElasticDiT offers a versatile, hardware-adaptive solution that eliminates the need for multiple specialized models, providing a promising path for future high-resolution image generation on mobile devices.
Fonte: arXiv cs.CV
Theory/Optimization • Score 85
Position: Ideas Should be the Center of Machine Learning Research
arXiv:2605.15253v1 Announce Type: new
Abstract: Machine learning research increasingly bifurcates into two disconnected modes: benchmark-driven engineering that prioritizes metrics over understanding, and idealized theory that often fails to transfer to modern systems. In this position paper, we argue that the field focuses too heavily on these endpoints, neglecting the central scientific object: the idea. We propose an Ideas First framework in which ideas are valued for the behavioral signatures they predict in modern models, and these signatures are tested through tailored experiments designed to detect the relevant patterns rather than to win leaderboards. This shift not only bridges the gap between theory and practice but also promotes equity by removing the "complexity premium," enabling rigorous scientific contributions from researchers with modest computational, financial, and human resources. Ultimately, we advocate for a research culture centered on ideas, treating benchmarks and theorems as instruments for testing mechanistic hypotheses rather than as ends in themselves.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Optimizing LLM Inference: Fluid-Guided Online Scheduling with Memory Constraints
arXiv:2504.11320v3 Announce Type: replace-cross
Abstract: Large language models now serve millions of users daily, with providers incurring costs exceeding $700,000 per day. Each request requires token-by-token inference, making GPU scheduling central to latency, capacity, and cost. The difficulty is endogenous memory growth: generated tokens expand the Key-Value (KV) cache, and overflow can evict in-progress requests and waste prior computation. We formulate inference as a multi-stage online scheduling problem with endogenous memory growth, linear iteration times, and GPU-resident KV-cache constraints. We introduce a fluid model that characterizes equilibrium batch composition, memory requirement, and stability region. Guided by the fluid model, we design WAIT (Waiting for Accumulated Inference Threshold), a threshold-based admission rule for known output lengths, and Nested WAIT, which extends the rule to unknown output lengths by regulating how requests advance across decode-stage segments. Both algorithms approximate the fluid benchmark asymptotically under the stated memory conditions. Nested WAIT uses an additional safety buffer of moderate scale to hedge against memory-overflow-induced evictions under unknown output lengths. In Vidur simulations configured for Llama-2-7B on an A100 GPU, with supplemental real-GPU validation reported in the appendix, the policies enlarge the empirically observed stable operating range relative to widely used baseline algorithms and reduce latency especially in near-overloaded and overloaded regimes.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Painless Activation Steering: An Automated, Lightweight Approach for Post-Training Large Language Models
arXiv:2509.22739v3 Announce Type: replace-cross
Abstract: Language models (LMs) are typically post-trained for desired capabilities and behaviors via weight-based or prompt-based steering, but the former is time-consuming and expensive, and the latter is not precisely controllable and often requires manual trial-and-error. While activation steering (AS) promises a cheap, fast, and controllable alternative to the two existing post-training methods, current AS techniques require hand-crafted prompt pairs or labor-intensive feature annotation, making them more inconvenient than the plug-and-play methods such as Reinforcement Learning (RL) and Supervised Fine-Tuning (SFT). We introduce Painless Activation Steering (PAS), a family of fully automated methods that make AS readily usable with any given labeled dataset, with no need for prompt construction, feature labeling, or human intervention. We evaluate PAS on three open-weight models (Llama3.1-8B-Instruct, DeepSeek-R1-Distill-8B, and Nous-Hermes-2) and 18 tasks; we find that PAS reliably improves performance for behavior tasks, but not for intelligence-oriented tasks. The introspective variant (iPAS) delivers the strongest causal steering effects (10.1% on Bias, 5.2% on Morality, and 34.8% on Alignment). We also show PAS delivers additional gains on top of In-Context Learning (ICL) and SFT. PAS constructs a fast, lightweight activation vector that can be cheaply trained, easily stored, and activated at will. Our results provide a characterization of where AS helps, where it fails, and how to deploy it as a practical, automated LM post-training option.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Njord: A Probabilistic Graph Neural Network for Ensemble Ocean Forecasting
arXiv:2605.15470v1 Announce Type: new
Abstract: Ocean dynamics are inherently chaotic, yet existing machine learning ocean models produce only deterministic forecasts. We introduce Njord, a probabilistic data-driven model for ocean forecasting, applicable to both global and regional domains. Njord combines a deep latent variable framework with a graph neural network architecture, enabling sampling each forecast step in a single forward pass. We apply Njord globally at 0.25{\deg} resolution and regionally to the Baltic Sea at 2 km resolution. To scale to these large ocean grids we introduce K-means cluster meshes that adapt to irregular sea surface geometry. Experiments demonstrate strong performance on both domains compared to deterministic machine learning baselines, while also providing uncertainty estimates from the sampled ensemble forecasts. On the global OceanBench benchmark, Njord achieves the lowest errors on average across upper-ocean variables when evaluated against real-world observations, with the largest improvements in surface temperature prediction.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Polynomial Neural Sheaf Diffusion: A Spectral Filtering Approach on Cellular Sheaves
arXiv:2512.00242v3 Announce Type: replace-cross
Abstract: Sheaf Neural Networks equip graph structures with a cellular sheaf: a geometric structure which assigns local vector spaces (stalks) and a linear learnable restriction/transport maps to nodes and edges, yielding an edge-aware inductive bias that handles heterophily and limits oversmoothing. However, common Neural Sheaf Diffusion implementations rely on SVD-based sheaf normalization and dense per-edge restriction maps, which scale with stalk dimension, require frequent Laplacian rebuilds, and yield brittle gradients. To address these limitations, we introduce Polynomial Neural Sheaf Diffusion (PolyNSD), a new sheaf diffusion approach whose propagation operator is a degree-K polynomial in a normalised sheaf Laplacian, evaluated via a stable three-term recurrence on a spectrally rescaled operator. This provides an explicit K-hop receptive field in a single layer (independently of the stalk dimension), with a trainable spectral response obtained as a convex mixture of K+1 orthogonal polynomial basis responses. PolyNSD enforces stability via convex mixtures, spectral rescaling, and residual/gated paths, reaching new state-of-the-art results on both homophilic and heterophilic benchmarks, inverting the Neural Sheaf Diffusion trend by obtaining these results with just diagonal restriction maps, decoupling performance from large stalk dimension, while reducing runtime and memory requirements.
Fonte: arXiv stat.ML
Theory/Optimization • Score 85
Breakeven complexity: A new perspective on neural partial differential equation solvers
arXiv:2605.15399v1 Announce Type: new
Abstract: Neural surrogate solvers of partial differential equations (PDEs) promise dramatic speedups over numerical methods, especially in scenarios requiring many solves. However, current accuracy-based evaluations do not fully consider two central issues: (1) neural solvers incur substantial up-front costs for data generation, training, and tuning; and (2) classical solvers can also generate low-fidelity solutions at a sufficiently low simulation cost. To explicitly account for these realities and fully incorporate end-to-end costs, we propose an evaluation framework centered on breakeven complexity, a metric that counts the forward solves before a learned solver is cost-effective relative to an error-equivalent traditional solver. To evaluate this measure, we apply scaling laws to determine how much training budget to allocate to data generation and discuss how to achieve smooth error-matching in diverse settings. We evaluate the breakeven complexity of multiple neural PDE solvers on three PDEs on 2D periodic domains from APEBench and a novel benchmark of flows past multiple obstacles generated by the GPU-native PyFR code. Among other findings, our results suggest that neural PDE solvers become more effective as problems get harder in terms of cost, dimension, rollout, physics regime (e.g. higher Reynolds number), etc.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
PRB-RUPFormer: A Recursive Unified Probabilistic Transformer for Residual PRB Forecasting
arXiv:2605.15363v1 Announce Type: new
Abstract: Accurate forecasting of residual Physical Resource Blocks (PRBs) is critical for proactive network slice provisioning, energy-efficient operation, and spectrum-aware decision making in cellular systems, where residual PRBs serve as a practical proxy for short- and medium-term spectrum availability. Existing PRB prediction methods typically rely only on historical PRB values and are trained independently per carrier or sector, limiting their ability to capture cross-carrier dependencies and providing no measure of forecast uncertainty. Moreover, point forecasts alone are insufficient for robust spectrum-aware control under highly variable traffic conditions. This paper proposes PRB-RUPFormer, a recursive unified probabilistic Transformer for residual PRB forecasting. The proposed model jointly processes multivariate KPI time series using temporal, seasonal, and carrier-aware embeddings, preserving inter-metric temporal coupling during recursive rollout and stabilizing long-horizon forecasting. A single shared model is trained across all carriers and sectors of an eNB, enabling efficient learning of joint traffic dynamics with low computational overhead. Forecast uncertainty is captured through quantile-based prediction intervals, providing confidence-aware estimates of future PRB availability. Evaluations on six months of commercial LTE network data from multiple U.S. locations demonstrate median MAE below 0.05 and hit probabilities above 0.80 for both one-day and seven-day recursive forecasts. These probabilistic predictions directly support spectrum-aware RAN functions such as dynamic carrier activation, congestion avoidance, and proactive spectrum sharing, making the proposed framework well-suited for dynamic spectrum access scenarios.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
CAX-Agent: A Lightweight Agent Harness for Reliable APDL Automation
arXiv:2605.15218v1 Announce Type: new
Abstract: Large language models deployed for MAPDL finite-element simulation face practical reliability challenges: without structured execution control, tool encapsulation, and fault recovery, outputs may be inconsistent and task failures are common. The Agent Harness paradigm addresses this by inserting domain-specific orchestration middleware that manages tool lifecycles, workflow state, and recovery escalation. This paper presents the architecture of CAX-Agent, a lightweight agent harness purpose-built for MAPDL automation, and empirically evaluates one of its core components -- the recovery policy.CAX-Agent organizes execution into three layers -- LLM service, agent harness, and solver backend -- with a recovery ladder that escalates from deterministic rule patching through model-driven regeneration to context enrichment and human intervention. We evaluate three recovery strategies (no_recovery, rule_only, and model_only) on 50 standard structural benchmarks with three repeated runs per strategy (450 case-runs total). Two independent human raters score task completion under blind conditions; inter-rater agreement is strong (quadratic weighted Cohen's kappa = 0.84, 96 percent of score pairs within one point). Model_only achieves the best completion rate (0.9267), task score (3.59/4), total score (9.16/10), and zero-intervention rate (0.84), outperforming rule_only (0.7733, 3.17/4, 7.03/10, 0.00) and no_recovery (0.6933, 2.74/4, 5.60/10, 0.00) with large effect sizes (Cliff's delta = 0.81-0.87). The benchmark uses deliberately simple geometries to isolate recovery-policy effects; we discuss the scope of these findings and directions for broader validation.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
SkillSmith: Compiling Agent Skills into Boundary-Guided Runtime Interfaces
arXiv:2605.15215v1 Announce Type: new
Abstract: Recently, skills have been widely adopted in large language model (LLM)-based agent systems across various domains. In existing frameworks, skills are typically injected into the agent reasoning loop as contextual guidance once matched to a runtime task, enabling specialized task-solving capabilities. We find that this execution paradigm introduces two major sources of redundancy: irrelevant context injection and repeated skill-specific reasoning and planning. To this end, we propose SkillSmith, a boundary-first compiler-runtime framework that compiles skill packages offline into minimal executable interfaces. By extracting fine-grained operational boundaries from skills, SkillSmith enables agents to dynamically access and execute only the relevant components at runtime, thereby minimizing unnecessary context injection and redundant reasoning overhead. In the evaluation on SkillsBench benchmark, SkillSmith reduces solve-stage token usage by 57.44%, thinking iterations by 42.99%, solve time by 50.57% (2.02x faster), and token-proportional monetary cost by 57.44% compared with using raw-skills. Moreover, compiled artifacts produced by a stronger model can be reused by a smaller or more efficient runtime model, improving task accuracy in cases where raw skill interpretation fails. The source code and data are available at https://github.com/AetherHeart-AI/Aeloon.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Federated Learning of Spiking Neural Networks under Heterogeneous Temporal Resolutions
arXiv:2605.15355v1 Announce Type: new
Abstract: Spiking neural networks (SNNs) are biologically inspired energy-efficient models that use sparse binary spike-based communication between neurons, making them attractive for resource-constrained edge devices. Federated learning enables such devices to train collaboratively without sharing raw data. In time-series applications, edge devices often collect data at different time resolutions due to hardware and energy constraints. This temporal heterogeneity poses a fundamental challenge for federated learning: parameters learned at one temporal resolution do not necessarily transfer directly to another, which might result in the naive federated averaging being ineffective. Targeting SNNs and, more broadly, deep networks with stateful neurons, we propose a federated learning framework that addresses this temporal resolution mismatch. We investigate how neuron parameters learned from data at different temporal resolutions and model aggregation should be integrated. We evaluate the proposed framework across two SNN-native benchmark datasets (SHD and DVS-Gesture) under a range of resolution heterogeneity scenarios. Our results show that the proposed adaptation methods can substantially recover accuracy lost due to temporal mismatch, hence enabling each client to train at their local temporal resolution while remaining compatible with the global model.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Tadpole: Autoencoders as Foundation Models for 3D PDEs with Online Learning
arXiv:2605.15284v1 Announce Type: new
Abstract: We introduce Tadpole, a novel foundation model for three-dimensional partial differential equations (PDEs) that addresses key challenges in transferability, scalability to high dimensionality, and multi-functionality. Tadpole is pre-trained as an autoencoder on synthetic 3D PDE data generated by an efficient online data-generation framework. This enables large-scale, diverse training without storage or I/O overhead, demonstrated by scaling to an equivalent of hundreds of terabytes of training data. By autoencoding single-channel spatial crops, Tadpole learns rich and transferable representations across heterogeneous physical systems with varying numbers of state variables and spatial resolutions. Although pre-trained solely as an autoencoder, Tadpole can be efficiently applied for multiple downstream tasks beyond reconstruction, including dynamics learning and generative modeling. For dynamics learning, we propose a novel parameter-efficient fine-tuning strategy that integrates low-rank adaptation, latent-space transformations, and reintroduced skip connections, achieving accurate temporal modeling with a minimal number of trainable parameters. Tadpole demonstrates strong fine-tuning performance across various downstream tasks, highlighting its versatility and effectiveness as a foundation model for 3D PDE learning. Source code and pre-trained weights of Tadpole are available at https://github.com/tum-pbs/tadpole
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
ScreenSearch: Uncertainty-Aware OS Exploration
arXiv:2605.16024v1 Announce Type: new
Abstract: Desktop GUI agents operate under partial observability: visually similar screens can correspond to different underlying workflow states, so locally plausible actions can lead to sharply different outcomes. We frame this as a problem of computer/OS state exploration, where effective behavior requires both expanding the reachable frontier and reducing ambiguity before committing. We present ScreenSearch, a system that combines structural screen retrieval and deduplication with an ambiguity-aware PUCT graph-bandit for large-scale desktop exploration. The retrieval layer converts UIA trees into location-aware structural features, indexes related screens through sparse token search and metadata filters, and maintains a shared deduplicated state graph across VM workers. On top of this graph, we define a scalable ambiguity signal based on matched-action outcome dispersion. If similar screens produce different next states under the same action signature, the state should be probed further rather than treated as resolved. We use this signal together with frontier rewards to drive large-scale exploration and replay-start policy evaluation over the shared graph. Across 11 desktop applications, ScreenSearch collects over 1M screenshots and over 30K deduplicated states, yielding large exploration corpora with substantial cross-application and within-application diversity. On a fixed replay-start slice, we observe a clear novelty--ambiguity trade-off: some policies reduce ambiguity quickly while discovering little frontier. Ambiguity reduction alone is therefore not a sufficient exploration objective. Appendix ablations show that stronger proposal priors can materially improve unique-state discovery during corpus building. These results suggest that state identity, proposal quality, and ambiguity-aware search all matter when deciding when to probe and when to commit.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
STAR: A Stage-attributed Triage and Repair framework for RCA Agents in Microservices
arXiv:2605.15581v1 Announce Type: new
Abstract: LLM-based root cause analysis (RCA) agents have recently emerged as a promising paradigm for incident diagnosis in microservice AIOps. However, their reliability remains fragile: an error in early evidence collection, hypothesis formulation, or causal analysis can propagate through the reasoning trace and eventually corrupt the final diagnosis. In this paper, we present \textbf{STAR}, a \emph{Stage-attributed Triage and Repair} framework for repairing erroneous RCA traces. STAR explicitly decomposes an RCA workflow into four structured stages, namely \emph{Evidence Package} (EP), \emph{Hypothesis Set} (HS), \emph{Analysis Structure} (AS), and \emph{Decision Report} (DR), and treats agent failure as a stage-localizable reasoning bug rather than a monolithic end-to-end error. Built on top of LangGraph, STAR performs stage-wise auditing, budget-aware \emph{Fast/Slow Routing}, \emph{decisive stage localization via counterfactual candidate evaluation}, and stage-specific patch-and-replay repair.
We evaluate STAR on a public large-scale benchmark and a real-world production dataset, using two RCA agent workflows and three foundation models. Experimental results show that STAR consistently improves both root cause localization and fault type classification over strong baselines. Moreover, STAR identifies the decisive faulty stage with high accuracy, repairs most initially incorrect traces within one or two replay rounds, and benefits substantially from both Fast/Slow Routing and counterfactual stage evaluation. These results suggest that explicitly modeling \emph{where} an RCA agent fails is an effective path toward reliable, debuggable, and self-repairing agentic RCA systems.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
SDOF: Taming the Alignment Tax in Multi-Agent Orchestration with State-Constrained Dispatch
arXiv:2605.15204v1 Announce Type: new
Abstract: Multi-agent orchestration frameworks such as LangChain, LangGraph, and CrewAI route tasks through graph-based pipelines but do not enforce the stage constraints that govern real business processes. We present SDOF, a framework that treats multi-agent execution as a constrained state machine. SDOF operates through two primary defensive layers, implemented by three components: (1) an Online-RLHF Specialized Intent Router trained via Generative Reward Modeling (GRPO) and (2) a StateAwareDispatcher with GoalStage finite-automaton checks and precondition/postcondition SkillRegistry validation for auditable execution control. On a recruitment system backed by the Beisen iTalent platform (6000+ enterprises), 185 expert-curated scenarios trigger 1671 live API calls. Our GSPO-aligned 7B Intent Router achieves higher joint accuracy than zero-shot GPT-4o on this FSM-constrained adversarial routing benchmark (80.9% versus 48.9%). In end-to-end execution, SDOF reaches 86.5% task completion (95% confidence interval 80.8 to 90.7) and blocks all 22 operations in the injection, illegal HR subset. Under a broader message-level blocking audit, SDOF attains precision 100% and recall 88%, expert agreement kappa=0.94. A separate evaluation on 960 SGD-derived dialogues spanning 8 service domains surfaces 201 stage-order conflicts under our FSM mapping, 41 of which arise in the normal split. This arXiv version reports the current validated scope; extended multi-seed training comparisons and deeper workflow evaluations will be released in a subsequent update.
Fonte: arXiv cs.AI
NLP/LLMs • Score 95
PRISM: Prompt Reliability via Iterative Simulation and Monitoring for Enterprise Conversational AI
arXiv:2605.15665v1 Announce Type: new
Abstract: Deploying large language model (LLM)-driven conversational agents in enterprise settings requires prompts that are simultaneously correct at launch and resilient to the non-deterministic behavioral drift that characterizes production LLM deployments. Existing prompt optimization frameworks address prompt quality as a one-time compile-time problem, leaving open the equally critical question of how to detect and repair prompt regressions caused by silent LLM behavior changes over time. We present PRISM (Prompt Reliability via Iterative Simulation and Monitoring), a closed-loop framework that treats prompt engineering as a continuous reliability engineering problem rather than a one-time authorship task. PRISM takes as input plain-language agent requirements, a set of configured tools and memory variables, and an initial draft prompt. It automatically generates test cases from requirements, simulates full multi-turn conversations against a platform-faithful LLM environment, evaluates pass/fail using an LLM-as-judge, diagnoses root causes of failures, and surgically repairs the prompt -- iterating until all tests pass. Critically, PRISM is designed to run on a scheduled basis (daily), treating LLM behavioral drift as a first-class reliability concern. We evaluate PRISM across 35 enterprise conversational agents over a three-week deployment period on the Yellow.ai V3 platform. PRISM reduces median prompt authoring time from 2 days to under 30 minutes, achieves 99% production reliability across all evaluated agents, and successfully identifies and repairs production regressions caused by LLM behavioral drift within a 24-hour detection window. Our results suggest that continuous, simulation-driven prompt optimization is both tractable and necessary for reliable enterprise conversational AI at scale.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
Preconditioned Regularized Wasserstein Proximal Sampling
arXiv:2509.01685v2 Announce Type: replace
Abstract: We consider sampling from a Gibbs distribution by evolving finitely many particles. We propose a preconditioned version of a recently proposed noise-free sampling method, governed by approximating the score function with the numerically tractable score of a regularized Wasserstein proximal operator. This is derived by a Cole--Hopf transformation on coupled anisotropic heat equations, yielding a kernel formulation for the preconditioned regularized Wasserstein proximal. The diffusion component of the proposed method is also interpreted as a modified self-attention block, as in transformer architectures. For quadratic potentials, we provide a discrete-time non-asymptotic convergence analysis and explicitly characterize the bias, which is dependent on regularization and independent of step-size. Experiments demonstrate acceleration and particle-level stability on various log-concave and non-log-concave toy examples to Bayesian total-variation regularized image deconvolution, and competitive/better performance on non-convex Bayesian neural network training when utilizing variable preconditioning matrices.
Fonte: arXiv stat.ML
Theory/Optimization • Score 85
MaxSketch: Robust Distinct Counting in Streams via Random Projections
arXiv:2605.15571v1 Announce Type: new
Abstract: Estimating the number of distinct elements in a data stream is well understood when repeated elements are identical. In modern settings, however, observations are high-dimensional and noisy, so repeated instances of the same object are only approximately similar -- for example, different images of the same individual may vary significantly at the pixel level. Classical sketches such as HyperLogLog rely on consistent hash values for identical elements and break down in this regime. Recent work on robust distinct counting in general metric spaces achieves $\widetilde{\Theta}(\sqrt{n})$ memory, which is tight in the worst case. We show that substantially improved memory guarantees are possible under geometric structure common in learned representations. We introduce MaxSketch, a simple max-linear sketch built from random Gaussian projections, and prove that it succeeds in estimating the number of distinct latent objects. Concretely, we show that under this assumption $m = \widetilde{O} (\log n / \varepsilon^2)$ random projections (and hence $\widetilde{O} (\log n/\varepsilon^2)$ memory) suffice to recover the true distinct count within a $(1+\varepsilon)$ factor. Experiments on image streams confirm that MaxSketch accurately estimates distinct counts and generalizes beyond the training regime. Our results bridge classical streaming algorithms and modern representation learning, showing how geometric structure can fundamentally reduce the complexity of distinct counting.
Fonte: arXiv stat.ML
RL • Score 85
GRLO: Towards Generalizable Reinforcement Learning in Open-Ended Environments from Zero
arXiv:2605.15464v1 Announce Type: new
Abstract: Post-training has become a crucial step for unlocking the capabilities of large language models, with reinforcement learning (RL) emerging as a critical paradigm. Recent RL-based post-training has increasingly split into two paradigms: reinforcement learning from human feedback (RLHF), which optimizes models using human preference signals in target domains, and reinforcement learning from verifiable rewards (RLVR), which operates in verifier-backed environments. The latter has dominated recent reasoning-oriented post-training because it delivers stronger gains and higher efficiency on domain-specific tasks (e.g., reasoning). However, although in-domain RL training achieves promising performance, it still requires a substantial amount of GPU compute, which remains a major barrier to broad adoption. In this work, we study the generalization ability of RLHF learned from scratch from a small set of interactions in open-ended environments, and investigate whether the conversational abilities it explicitly acquires can implicitly transfer to downstream tasks such as mathematical reasoning and code generation, namely GRLO. Specifically, on Qwen3-4B-Base backbone, GRLO improves the average performance across all domains from 24.1 to 63.1 with only 5K prompts and 22.7 GPU hours, requiring about $46\times$ less data and $68\times$ less compute than a strong in-domain RLVR baseline. The resulting model is even competitive with Qwen's released post-trained models which required a much larger training cost. Notably, a subsequent in-domain RLVR stage brings only selective gains, mainly on harder competition-math benchmarks. We hope GRLO offers a simple and efficient recipe for building broadly capable post-trained models. Our code and data will be available at: \href{https://github.com/SJY8460/GRLO}{https://github.com/SJY8460/GRLO}.
Fonte: arXiv cs.LG
RL • Score 85
DualKV: Shared-Prompt Flash Attention for Efficient RL Training with Large Rollouts and Long Contexts
arXiv:2605.15422v1 Announce Type: new
Abstract: Modern RL post-training methods such as GRPO and DAPO train on $N$ response sequences of $R$ tokens sampled from a shared prompt of $P$ tokens, but standard FlashAttention replicates all $P$ prompt tokens $N$ times across both forward and backward passes -- duplicating compute and memory on identical hidden states. In large-rollout, long-context RL training ($N{\geq}16$, $P{\geq}8\text{K}$), this redundancy dominates the policy update cost. We observe that in decoder-only models, causal masking makes prompt representations invariant across sequences at every layer, so all per-token operations (norms, projections, MLP) and attention can process the prompt once -- a property not yet exploited at the kernel level for training. We propose \textbf{DualKV}, the first FlashAttention kernel variant that eliminates shared-prompt replication during RL training, via (1)~fused CUDA forward and backward kernels that iterate over two disjoint KV regions -- shared context and per-sequence response -- in a single kernel launch, and (2)~a data-pipeline redesign in veRL that repacks $N(P{+}R)$ tokens into $P{+}NR$ tokens per micro-batch, extending the token reduction from attention to the entire model by a factor $\rho = N(P{+}R)/(P{+}NR)$. DualKV is mathematically equivalent to standard attention and introduces no approximation. On Qwen3-8B GRPO training with 8$\times$H100 GPUs ($N{=}32$, 8K-context), DualKV achieves $1.63$--$2.09\times$ policy-update speedup, enables $2\times$ larger micro-batches, and raises MFU from $36\%$ to $76\%$. Similar gains hold for DAPO ($2.47\times$ speedup, $77\%$ MFU). At 30B MoE scale on 16$\times$H100, DualKV achieves $3.82\times$ policy-update and $3.38\times$ end-to-end step speedup over FlashAttention (which requires 4-way Ulysses sequence parallelism to avoid OOM).
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Layer-wise Derivative Controlled Networks
arXiv:2605.15463v1 Announce Type: new
Abstract: As machine learning models grow in complexity, they increasingly struggle with three conflicting demands: the need for high accuracy, the requirement for hardware efficiency, and the necessity of functional stability. Traditional architectures often achieve performance at the expense of spiky or unpredictable behavior, where small changes in input lead to massive swings in output -- a critical flaw for real-world deployment in sensitive environments. This paper introduces ChainzRule (CR), a novel neural architecture designed to harmonize these competing goals. ChainzRule replaces standard piecewise-linear activations with a Polynomial Engine governed by Differential Regularization (DREG). Unlike traditional methods that impose global, coarse-grained constraints on a model's Lipschitz constant, DREG acts as a targeted regularization on intermediate derivatives. This approach suppresses extreme sensitivity without attenuating the representational power inherent in the Polynomial Engine. In head-to-head "Fair Fight" benchmarks, ChainzRule outperformed standard models while using 15.5x fewer parameters. On the MNIST dataset, it reduced peak gradient volatility by an average of 23.1%, ensuring a smoother and more predictable manifold. On Yelp Full ordinal regression under explicit DREG regularization, ChainzRule achieves 70.17% accuracy, validating that derivative-aware regularization is compatible with competitive performance on realistic tasks. By embedding gradient awareness into the architecture via DREG, ChainzRule demonstrates that stability and accuracy need not be competing objectives.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
On the Stability of Growth in Structural Plasticity
arXiv:2605.15435v1 Announce Type: new
Abstract: Standard deep-learning pipelines usually choose the network architecture before training and keep it fixed throughout optimization. In contrast, a model can also be adapted by editing its structure during training, for example by pruning existing hidden-neuron units or growing new ones. Although growth is appealing for adaptive and continual systems, we show that it is not simply the inverse of pruning. Pruning selects among units that have participated in training from the start, whereas growth inserts new units into an already specialized optimization trajectory. We isolate this insertion problem and show that newborn units are often forward-active but backward-starved: they participate in the forward computation, yet receive much weaker gradient signal than incumbent units. This disadvantage is minor in small MLP benchmarks, but becomes clear in harder image-classification settings with a convolutional trunk. In these settings, \textsc{Grow} can achieve high final accuracy during the structural-editing procedure, while \textsc{Prune} is stronger when performance is averaged over the training trajectory or when the final sparse network is retrained from scratch. Interventions targeting optimizer state, insertion, selection, and trainability show that improving the integration of newborn units can improve adaptive performance, but does not automatically produce better final subnetworks. In continual-learning benchmarks stressing plasticity loss, \textsc{Grow} becomes competitive mainly when new units have enough time to integrate. Together, these results suggest that \textsc{Grow} should be evaluated not only as an architecture-search operator, but as a time-sensitive optimization process whose success depends on insertion stability.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Time-Varying Deep State Space Models for Sequences with Switching Dynamics
arXiv:2605.15311v1 Announce Type: new
Abstract: The identification and modeling of time-varying systems is a fundamental challenge in signal processing and system identification. To address this challenge, we propose a class of time-varying state-space model (SSM) based neural networks in which the neurons' states are governed by time-varying dynamics. The proposed model provides the learnable time-varying dynamics through a dictionary of basis functions, where each basis function evolves differently over time. We evaluate the proposed approach on both synthetic data from switching systems and a speech denoising task where real audio is corrupted with switching dynamics noise. The results show that the proposed time-varying model consistently outperforms its time-invariant counterparts while maintaining comparable computational complexity. Our investigations also reveal which aspects of the time-varying dynamics of the data most need to be captured by the proposed time-invariant models, how the additional freedom provided by time-varying basis functions should be allocated across model components, and to what extent larger models can compensate for time-invariant limitations.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
$f$-Trajectory Balance: A Loss Family for Tuning GFlowNets, Generative Models, and LLMs with Off- and On-Policy Data
arXiv:2605.15417v1 Announce Type: new
Abstract: In GFlowNets and variational inference, it has been shown that the mean square error between target and model log probabilities is an effective, low variance, surrogate loss for training generative models.
This loss has the property that when evaluated \emph{on-policy} its gradients correspond to those of the KL divergence, while \emph{off-policy} it remains a valid loss with the same global minimizer. In this work, we demonstrate that this construction can be extended to the whole family of $f$-divergences, leading to a family of losses whose on-policy gradients are that of the corresponding $f$-divergence, but retain the same global minimizer off-policy. Specifically, we show that the on-policy gradients lead to a one to one correspondence between translation invariant loss functions on the target and model log probabilities, and $f$-divergences. This equivalence allows us to design new surrogate loss functions for tuning a wide class of generative models that inherit the properties of the corresponding $f$-divergence, such as being more mode covering, whilst being applicable to off-policy data. We apply our losses on a range of tasks, including classic synthetic examples, SynFlowNets for molecule discovery, and asynchronous large language model (LLM) tuning, demonstrating that our models retain their predicted properties on- and off-policy in a wide class of generative models.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
NIMO Controller: a self-driving laboratory orchestrator based on the Model Context Protocol
arXiv:2605.15227v1 Announce Type: new
Abstract: Self-driving laboratories (SDLs) have attracted increasing attention as a means of accelerating scientific discovery; however, developing SDL software remains technically demanding. To improve accessibility, orchestration software frameworks have been proposed to coordinate SDL components. Nevertheless, existing frameworks are primarily designed for human interaction and do not provide standardized interfaces suitable for AI agents. In this work, we propose an SDL software architecture based on the Model Context Protocol (MCP), in which all SDL functionalities are exposed through MCP servers. Following this design principle, we introduce an MCP-based SDL orchestrator, named NIMO Controller. It provides a visual programming interface automatically generated through MCP-based tool discovery, allowing human users to design experimental workflows without writing code. The same MCP backend can also be accessed by AI agents, providing a unified interface for both human users and AI agents. We demonstrate the proposed system through a case study on a color-matching SDL. The results validate the usability of the proposed MCP-based SDL architecture.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
Proposal-Guided Greedy Surrogate Refinement for PDE-Driven High-Dimensional Rare-Event Estimation
arXiv:2605.15356v1 Announce Type: cross
Abstract: Accurate surrogate construction for PDE-driven high-dimensional rare-event simulation is challenging when performance evaluations are expensive. Since a globally accurate surrogate may require many high-fidelity evaluations, adaptive importance sampling provides a natural localization tool: its evolving proposal distribution progressively identifies the failure-relevant region. Motivated by this observation, we propose a surrogate-assisted adaptive importance sampling framework that refines the surrogate locally along the evolving proposal, rather than over the entire input space. The surrogate combines an encoder with a neural network, providing a low-dimensional latent representation for both prediction and sample selection. At each adaptive iteration, candidates drawn from the current proposal are selected by a greedy latent-space rule balancing proximity to the estimated failure boundary and sample diversity. The selected samples are evaluated by the high-fidelity model and used to refine the surrogate, which then guides the subsequent cross-entropy-type adaptive proposal update. We establish one-step proposal stability bounds under local surrogate errors, together with surrogate-induced misclassification and finite-sample estimation error bounds. Numerical experiments on multimodal benchmarks and PDE-driven rare-event problems up to 100 dimensions show that the proposed method achieves accuracy comparable to true-model adaptive importance sampling while requiring substantially fewer high-fidelity evaluations.
Fonte: arXiv stat.ML
Multimodal • Score 85
LRCP: Low-Rank Compressibility Guided Visual Token Pruning for Efficient LVLMs
arXiv:2605.15621v1 Announce Type: new
Abstract: Large vision-language models (LVLMs) achieve strong multimodal understanding, but their inference cost grows rapidly with the number of visual tokens, especially for high-resolution images and long videos. Existing attention-based methods estimate token importance from attention scores, which may introduce positional bias, while representation-based methods reduce visual redundancy based on feature relations or reconstruction errors, overlooking the global structure of the visual token set. In this paper, we revisit visual token compression from the perspective of low-rank compressibility. Across models and datasets, we observe that visual token representations exhibit a pronounced low-rank structure, with a dominant subspace that remains stable even after a large fraction of tokens is randomly removed. Motivated by this finding, we propose LRCP, a training-free compression framework that first estimates the dominant low-rank subspace of visual tokens via PCA, and then scores each token by its projection residual onto this subspace, retaining tokens that are poorly explained by the low-rank background. Extensive experiments show that LRCP achieves superior results, preserving 94.7% of the original image-understanding performance with an 88.9% token reduction and 97.8% of the average video-understanding accuracy with an 87.5% token reduction.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Always Learning, Always Mixing: Efficient and Simple Data Mixing All The Time
arXiv:2605.15220v1 Announce Type: new
Abstract: Data mixing decides how to combine different sources or types of data and is a consequential problem throughout language model training. In pretraining, data composition is a key determinant of model quality; in continual learning and adaptation, it governs what is retained and acquired. Yet existing data mixing methods address only one phase of this lifecycle at a time: some require smaller proxy models tied to a single training phase, others assume a fixed domain set, and continual learning lacks principled guidance altogether. We argue that data mixing is fundamentally an online decision making problem -- one that recurs throughout training and demands a single, unified solution. We introduce OP-Mix (On-Policy Mix), a data mixing algorithm that operates across the entire language model training lifecycle. Our main insight is that candidate data mixtures can be cheaply simulated by interpolating between low-rank adapters trained directly on the current model, eliminating separate proxy models and ensuring the search is always grounded in the model's actual learning dynamics. Across pretraining, continual midtraining, and continual instruction tuning, OP-Mix consistently finds near-optimal mixtures while using a fraction of the compute of the baselines. In pretraining, OP-Mix improves upon training without mixing by 6.3% in average perplexity. For continual learning, OP-Mix matches the performance of both retraining and on-policy distillation while using 66% and 95% less overall compute, respectively. OP-Mix suggests a different view of language model training: not a sequence of distinct phases, but a single continuous process of learning from data.
Fonte: arXiv cs.CL
MLOps/Systems • Score 85
FRESH: Information-Geometric Calibration of Patient-Level Models to Aggregate Evidence
arXiv:2605.16246v1 Announce Type: cross
Abstract: This note introduces FRESH (Fusion of Recent Evidence and Subject Histories), a method for incorporating population-level summary results -- published clinical trials, registry summaries,
prior natural-history studies, and peer-reviewed indirect comparisons -- into predictive models trained on patient-level data. This method provides a principled means of combining both
patient-level and aggregate-level data types into a unified data-efficient model for clinical decision making.
FRESH assumes access to a generative model trained on patient-level data sources (e.g. clinical trial or real-world data). The method produces patient-level predictions from a re-calibrated
model that matches a set of specified aggregate statistics for a target population. This can be understood as a patient-level recapitulation of the aggregate source -- with the key property
that the recalibration is a minimal perturbation of the original joint distribution in a specific information-geometric sense. The resulting samples can be analyzed directly or combined into a
post-training procedure to update the original generative model.
This approach enables several applications where rigorously incorporating patient-level data with summary information is valuable, including (i) contextualizing single-arm trial results with
respect to recent standard-of-care, (ii) clinical-trial simulations for design and probability-of-technical-success estimation, and (iii) comparative-effectiveness analyses of on-market
therapies.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
ALSO: Adversarial Online Strategy Optimization for Social Agents
arXiv:2605.15768v1 Announce Type: new
Abstract: Social simulation provides a compelling testbed for studying social intelligence, where agents interact through multi-turn dialogues under evolving contexts and strategically adapting opponents. Such environments are inherently non-stationary, requiring agents to dynamically adjust their strategies over time. However, most Large Language Model (LLM) based social agents rely on static personas, while existing approaches for enhancing social intelligence, such as offline reinforcement learning or external planners, are ill-suited to these settings, typically assuming stationarity and incurring substantial training overhead. To bridge this gap, we propose \textbf{ALSO} (\textbf{A}dversarial on\textbf{L}ine \textbf{S}trategy \textbf{O}ptimization), the first framework for online strategy optimization in multi-agent social simulation. ALSO advances social adaptation through two key contributions. (1) ALSO formulates multi-turn interaction as an adversarial bandit problem, where combinations of static personas and dynamic strategy instructions are treated as arms, providing a principled solution to non-stationarity without relying on environmental stability assumptions. (2) To predict rewards and generalize sparse feedback in multi-turn dialogues, ALSO introduces a lightweight neural surrogate to predict rewards from interaction histories, enabling sample-efficient exploration and continuous online adaptation. Experiments on the Sotopia benchmark demonstrate that ALSO consistently outperforms static baselines and existing optimization methods in dynamic environments, validating the effectiveness of adversarial online strategy optimization for building robust social agents.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Ensemble Monitoring for AI Control: Diverse Signals Outweigh More Compute
arXiv:2605.15377v1 Announce Type: new
Abstract: As AI systems are increasingly deployed in autonomous agentic settings at scale, it is important to ensure the actions they take are safe and aligned with user intent. Monitoring agent actions is a key safety mechanism, yet reliable monitors remain difficult to build and the scale of these systems makes human oversight impractical. We show that combining signals from diverse monitors into an ensemble improves detection of misaligned actions. We build 12 GPT-4.1-Mini monitors using both prompting and fine-tuning strategies. We evaluate them on coding tasks where candidate solutions pass standard tests but fail on adversarial inputs. In this setting, diverse ensembles outperform both individual monitors and homogeneous ensembles. Our best 3-monitor ensemble achieves 2.4x greater detection performance gain compared to an ensemble composed of three identical monitors, with the same ensemble performing strongly on an independent dataset. We contend that these results show that diversity - not scale - drives gains. The best ensembles combine strong individual performance with low correlation between monitors. Furthermore, fine-tuned monitors appear in every top-performing ensemble and maintain this advantage on out-of-distribution attack types, suggesting that fine-tuning enables detection capabilities that prompting alone does not elicit. These results support ensemble monitoring as a practical AI control strategy for safety gains at reasonable inference costs.
Fonte: arXiv cs.AI
Vision • Score 85
MaTe: Images Are All You Need for Material Transfer via Diffusion Transformer
arXiv:2605.15660v1 Announce Type: new
Abstract: Recent diffusion-based methods for material transfer rely on image fine-tuning or complex architectures with assistive networks, but face challenges including text dependency, extra computational costs, and feature misalignment. To address these limitations, we propose MaTe, a streamlined diffusion framework that eliminates textual guidance and reference networks. MaTe integrates input images at the token level, enabling unified processing via multi-modal attention in a shared latent space. This design removes the need for additional adapters, ControlNet, inversion sampling, or model fine-tuning. Extensive experiments demonstrate that MaTe achieves high-quality material generation under a zero-shot, training-free paradigm. It outperforms state-of-the-art methods in both visual quality and efficiency while preserving precise detail alignment, significantly simplifying inference prerequisites.
Fonte: arXiv cs.CV
Theory/Optimization • Score 85
$\phi$-Balancing for Mixture-of-Experts Training
arXiv:2605.15403v1 Announce Type: new
Abstract: Mixture-of-Experts (MoE) models rely on balanced expert utilization to fully realize their scalability. However, existing load-balancing methods are largely heuristic and operate on noisy mini-batch assignment statistics, introducing bias relative to population-level objectives. We propose $\phi$-balancing, a principled framework that directly targets population-level expert balance by minimizing a strictly convex, symmetric, and differentiable potential of the expected routing distribution. Using convex duality, we derive an equivalent min-max formulation and obtain a simple online algorithm via mirror descent, yielding an efficient EMA-based routing adjustment with negligible overhead. Across large-scale pretraining and downstream fine-tuning, $\phi$-balancing consistently outperforms prior Switch-style and loss-free baselines, demonstrating more stable and effective expert utilization.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
GQLA: Group-Query Latent Attention for Hardware-Adaptive Large Language Model Decoding
arXiv:2605.15250v1 Announce Type: new
Abstract: Multi-head Latent Attention (MLA), the attention used in DeepSeek-V2/V3, jointly compresses keys and values into a low-rank latent and matches the H100 roofline almost perfectly. Its trained weights, however, expose only one decoding path - an absorbed MQA form - which ties efficient inference to H100-class compute-bandwidth ratios, forfeits tensor parallelism along the head axis, and yields no Multi-Token Prediction (MTP) gain on commodity inference GPUs such as the export-restricted H20. We propose Group-Query Latent Attention (GQLA), a minimal modification of MLA whose trained weights expose two algebraically equivalent decoding paths over the same parameters: an MQA-absorb path identical to MLA's, and a GQA path with a per-group expanded cache. The runtime picks the path that matches the target hardware - no retraining, no custom kernels - so a single set of GQLA weights pins the rooflines of both H100 (MQA-absorb, s_q=1) and H20 (GQA + MTP, s_q=2), while supporting up to 8-way zero-redundancy tensor parallelism on the GQA path. To avoid pretraining from scratch we extend TransMLA into TransGQLA, which converts a pretrained GQA checkpoint into a GQLA model; on LLaMA-3-8B it compresses the per-token KV cache to 28.125% of the GQA baseline on the MQA-absorb path while structurally preserving GQA-level traffic on the per-group path.
Fonte: arXiv cs.LG
Vision • Score 85
MR2-ByteTrack: CNN and Transformer-based Video Object Detection for AI-augmented Embedded Vision Sensor Nodes
arXiv:2605.15423v1 Announce Type: new
Abstract: Modern smart vision sensors need on-device intelligence to process video streams, as cloud computing is often impractical due to bandwidth, latency, and privacy constraints. However, these sensory systems typically rely on ultra-low-power microcontrollers (MCUs) with limited memory and compute, making conventional video object detection methods, which require feature storage or multi-frame buffering, unfeasible. To address this challenge, we introduce Multi-Resolution Rescored ByteTrack (MR2-ByteTrack), a Video Object Detection (VOD) method tailored for MCU-based embedded vision nodes. MR2-ByteTrack reduces computational cost by alternating between full- and low-resolution inference, while linking detections across frames via ByteTrack and correcting misclassifications through the Rescore algorithm, which applies probability union rules to aggregate detection confidence scores across frames. We apply our approach to both a CNN-based detector and a Transformer-based model, demonstrating its generality across architectures with fundamentally different spatial processing. Experiments on ImageNetVID demonstrate that MR2-ByteTrack maintains accuracy, achieving mAP scores of up to 49.0 for the CNN-based models and 48.7 for the Transformer, while reducing multiply-accumulate operations by as much as 53\% for the CNNs and 32\% for the Transformer. When deployed on GAP9, an ultra-low-power RISC-V multicore MCU, our method yields up to 55\% energy savings compared to processing only full-resolution images, enabling the first real-time Transformer-based VOD on an MCU-class embedded vision node. Code available at https://github.com/Bomps4/Multi_Resolution_Rescored_ByteTrack/tree/IEEE_Access
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
DimMem: Dimensional Structuring for Efficient Long-Term Agent Memory
arXiv:2605.15759v1 Announce Type: new
Abstract: Large language model (LLM) agents require long-term memory to leverage information from past interactions. However, existing memory systems often face a fidelity--efficiency trade-off: raw dialogue histories are expensive, while flat facts or summaries may discard the structure needed for precise recall. We propose \textbf{DimMem}, a lightweight dimensional memory framework that represents each memory as an atomic, typed, and self-contained unit with explicit fields such as time, location, reason, purpose, and keywords. This representation exposes the structure needed for dimension-aware retrieval, memory update, and selective assistant-context recall without storing full histories in the model context. Across LoCoMo-10 and LongMemEval-S, DimMem achieves \textbf{81.43\%} and \textbf{78.20\%} overall accuracy, respectively, outperforming existing lightweight memory systems while reducing LoCoMo per-query token cost by \textbf{24\%}. We further show that dimensional memory extraction is learnable by compact models: after fine-tuning on the DimMem schema, a Qwen3-4B extractor surpasses LightMem with GPT-4.1-mini on both benchmarks and reaches performance comparable to, or better than, much larger extractors in key settings. These results suggest that explicit dimensional structuring is an effective and efficient foundation for long-term memory in LLM agents. Code is available at https://github.com/ChowRunFa/DimMem.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
Measuring Maximum Activations in Open Large Language Models
arXiv:2605.15572v1 Announce Type: new
Abstract: The dynamic range of activations is a first-order constraint for low-bit quantization, activation scaling, and stable LLM inference. Prior work characterized outlier features and massive activations on pre-2024 LLaMA-style models, and the downstream activation-quantization stack inherits that picture without revisiting it for the post-LLaMA open-model boom. We ask the deployment-oriented question: how large can activations get in modern open LLMs, and how does this magnitude vary across families, generations, and training stages? Under a unified pipeline (5,000-sample multi-domain corpus, family-specific tokenization, identical hooks across embeddings, hidden states, attention, MLP/MoE, SwiGLU gates, and final norm), we measure global and layerwise maxima on 27 checkpoints from 8 open families spanning dense, MoE, vision-language, intermediate-training, and instruction-tuned variants. We find that (i) global maxima span over nearly four orders of magnitude at comparable parameter counts, with Qwen3.5 and MoE checkpoints in the 10^2 to 10^3 range and Gemma3-27B-it reaching ~7 x 10^5; (ii) cross-family and cross-generation comparisons break simple monotonic scaling; and (iii) MoE checkpoints exhibit 14.0-23.4x lower peaks than matched-scale dense counterparts, while the residual stream carries the global maximum in 22/24 checkpoints. A lightweight INT-8 sanity check shows that measured maxima co-vary with low-bit reconstruction error via activation-scale selection. We conclude that maximum activation magnitude is a model property tied to family, architecture, and training stage - not a simple byproduct of size - and should be measured and reported alongside any open-weight release before low-bit deployment. The code is publicly available at https://github.com/clx1415926/Max_act_llm.
Fonte: arXiv cs.CL
Theory/Optimization • Score 85
Curriculum Learning of Physics-Informed Neural Networks based on Spatial Correlation
arXiv:2605.15254v1 Announce Type: new
Abstract: Physics-Informed Neural Networks (PINNs) combine deep learning with physical constraints for solving partial differential equations (PDEs), and are widely applied in fluid mechanics, heat transfer, and solid mechanics. However, PINN training still suffers from high-dimensional non-convex loss landscapes, imbalanced multiobjective constraints, and ineffective information propagation. Existing curriculum learning and causality-guided strategies improve training stability, but mainly focus on temporal or parametric progression, lacking explicit treatment of spatial information propagation and inter-region consistency. Moreover, they are not directly applicable to boundary value problems (BVPs) with strong spatial coupling. To address this issue, we propose a spatially correlated curriculum learning framework for PINNs. To the best of our knowledge, this is the first work to address PINN training difficulties from the perspective of spatial coupling among subregions. First, spatial causal weights guide information from near-boundary regions inward, reducing optimization failures and spurious convergence. Second, a low-frequency information bridge enforces pseudo-label-based consistency across spatially separated regions, suppressing global low-frequency drift. Third, a region-adaptive reweighting strategy adjusts subregion losses to reduce local residuals and recover high-frequency details. Experiments on PDE benchmarks show that, under comparable computational cost, the proposed method alleviates training failures and improves solution accuracy. The code is available at https://github.com/pigofmomo/CurriculumLearningPINN.
Fonte: arXiv cs.LG
Multimodal • Score 85
DRS-GUI: Dynamic Region Search for Training-Free GUI Grounding
arXiv:2605.15542v1 Announce Type: new
Abstract: GUI agents powered by Multimodal Large Language Models (MLLMs) have demonstrated impressive capability in understanding and executing user instructions. However, accurately grounding instruction-relevant elements from high-resolution screenshots cluttered with irrelevant UI components remains challenging for existing approaches. Inspired by how humans dynamically adjust their perceptual scope to locate task-related regions on complex screens, we propose DRS-GUI, a training-free dynamic region search framework for GUI grounding that can be seamlessly integrated into existing MLLMs. DRS-GUI introduces a lightweight UI Perceptor that performs three human-like perceptual actions (Focus, Shift, and Scatter) to progressively explore the interface and generate region proposals. To dynamically schedule these actions, we further design an Action Planner based on Monte Carlo Tree Search (MCTS). A region quality reward is employed to evaluate and select the highly instruction-relevant region, efficiently pruning redundant UI elements. Experiments demonstrate that DRS-GUI yields a 14\% improvement on ScreenSpot-Pro for general and GUI-specific MLLMs (Qwen2.5-VL-7B and UGround-V1-7B), significantly enhancing grounding performance and generalization.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
ColPackAgent: Agent-Skill-Guided Hard-Particle Monte Carlo Workflows for Colloidal Packing
arXiv:2605.15625v1 Announce Type: new
Abstract: We introduce ColPackAgent, an agent framework that autonomously runs Monte Carlo simulations of colloidal packing through a Model Context Protocol (MCP) tool server and an agent skill, whether as a standalone agent or inside an existing agent system. By harnessing the MCP server and agent skill, ColPackAgent executes a structured workflow for colloidal packing simulations, which are central to studies of phase behavior, self-assembly, and materials design. Without dedicated simulation tools and workflow instructions, general-purpose Large Language Model (LLM) agents tend to describe such workflows rather than execute them reliably. The MCP server exposes a custom-built colpack Python package that wraps HOOMD-blue hard-particle Monte Carlo, and the skill encodes a four-stage workflow contract. ColPackAgent can carry out the workflow interactively with human feedback, autonomously from an end-to-end prompt, or as autoresearch following a provided program file. We demonstrate the system in different modes with several colloidal packing simulation examples such as cube particles in 3D, a binary system of disks and capsules in 2D, and the 2D hard-disk freezing transition using autoresearch. We also compare model performance on this workflow across a panel of LLMs with 17 stage-specific prompts. This benchmark provides a stage-level check of how reliably different models follow the setup, planning, and analysis workflow. Together, these results show that pairing a domain Python package with MCP tools and a portable agent skill provides a practical route for turning a simulation toolkit into an agent-assisted research workflow.
Fonte: arXiv cs.AI
NLP/LLMs • Score 90
Agentic Discovery of Neural Architectures: AIRA-Compose and AIRA-Design
arXiv:2605.15871v1 Announce Type: new
Abstract: Toward recursive self-improvement, we investigate LLM agents autonomously designing foundation models beyond standard Transformers. We introduce a dual-framework approach: AIRA-Compose for high-level architecture search, and AIRA-Design for low-level mechanistic implementation. AIRA-Compose uses 11 agents to explore fundamental computational primitives under a 24-hour budget. Agents evaluate million-parameter candidates, extrapolating top designs to 350M, 1B, and 3B scales. This yields 14 architectures across two families: AIRAformers (Transformer-based) and AIRAhybrids (Transformer-Mamba). Pre-trained at 1B scale, these consistently outperform Llama 3.2 and Composer-found baselines. On downstream tasks, AIRAformer-D and AIRAhybrid-D improve accuracy by 2.4% and 3.8% over Llama 3.2. Furthermore, AIRA-Compose finds models with highly efficient scaling frontiers: AIRAformer-C scales 54% and 71% faster than Llama 3.2 and Composer's best Transformer, while AIRAhybrid-C outscales Nemotron-2 by 23% and Composer's best hybrid by 37%. AIRA-Design tasks 20 agents with writing novel attention mechanisms for long-range dependencies and high-performing training scripts. On the Long Range Arena benchmark, agent-designed architectures reach within 2.3% and 2.6% of human state-of-the-art on document matching and text classification. On the Autoresearch benchmark, Greedy Opus 4.5 achieves 0.968 validation bits-per-byte under a fixed time budget, surpassing the published minimum. Together, these frameworks show AI agents can autonomously discover architectures and algorithmic optimizations matching or surpassing hand-designed baselines. This establishes a powerful paradigm for discovering next-generation foundation models, marking a clear step toward recursive self-improvement.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Mask-Morph Graph U-Net: A Generalisable Mesh-Based Surrogate for Crashworthiness Field Prediction under Large Geometric Variation
arXiv:2605.15231v1 Announce Type: new
Abstract: Nonlinear finite element crash simulations are accurate but computationally expensive, limiting their use in iterative design optimisation. Machine-learning surrogate models based on graph neural networks (GNNs) offer a faster alternative. Message-passing GNNs are widely used for mesh simulation, and their shared node and edge update functions are relatively generalisable across varying graph structures. By contrast, non-shareable edge-specific aggregation layers can capture nonlinear relationships more accurately but usually require fixed graph connectivity, which limits generalisability. This paper presents Mask-Morph Graph U-Net (MMGUNet), a practical approach to addressing the limitation of hierarchical Graph U-Net architectures that use edge-specific downsampling and upsampling layers. Fixed coarse graph connectivity is required for edge-specific layers. To retain this while improving spatial correspondence, the proposed method morphs the coarsened graph hierarchy to each input mesh using feature-aligned barycentric parameterisation before constructing cross-graph edges. It further applies node masking during supervised pretraining, followed by parameter-efficient fine-tuning in which high-parameter edge-specific layers are frozen. The proposed approach is evaluated in in-distribution, out-of-distribution, and cross-component transfer settings using mean Euclidean distance and maximum intrusion percentage error. Results show that coarse-graph morphing improves test accuracy relative to a fixed-coarse-graph baseline, while masked supervised pretraining reduces the train-test discrepancy and improves data efficiency during transfer. The proposed model also achieves lower prediction error compared with external baselines. These results demonstrate a practical route toward reusable, data-efficient mesh-based surrogate modelling for crashworthiness design exploration.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
AgentStop: Terminating Local AI Agents Early to Save Energy in Consumer Devices
arXiv:2605.15206v1 Announce Type: new
Abstract: Autonomous agents powered by large language models (LLMs) are increasingly used to automate complex, multi-step tasks such as coding or web-based question answering. While remote, cloud-based agents offer scalability and ease of deployment, they raise privacy concerns, depend on network connectivity, and incur recurring API costs. Deploying agents locally on user devices mitigates these issues by preserving data privacy and eliminating usage-based fees. However, agentic workflows are far more resource-intensive than typical LLM interactions. Iterative reasoning, tool use, and failure retries substantially increase token consumption, often expending significant compute without successfully completing tasks.
In this work, we investigate the time, token, and energy overhead of locally deployed LLM-based agents on consumer hardware. Our measurements show that agentic execution increases GPU power draw, temperature, and battery drain compared to single-inference workloads. To address this inefficiency, we introduce AgentStop, a lightweight efficiency supervisor that predicts and preemptively terminates trajectories unlikely to succeed. Leveraging low-cost execution signals, such as token-level log probabilities, AgentStop can reduce wasted energy by 15-20% with minimal impact on task performance (<5% utility drop) for challenging web-based question answering and coding benchmarks. These findings position predictive early termination as a practical mechanism for enabling sustainable, privacy-preserving LLM agents on user devices. Our project code and data are available at https://github.com/brave-experiments/AgentStop.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Mini-JEPA Foundation Model Fleet Enables Agentic Hydrologic Intelligence
arXiv:2605.14120v1 Announce Type: new
Abstract: Geospatial foundation models compress multispectral observations into dense embeddings increasingly used in natural-language environmental reasoning systems. A single planetary-scale model, e.g. Google AlphaEarth, handles broad characterization well but may compromise on specialized hydrologic signals. Such generalist models are also often inaccessible, expensive, and require large-scale compute. We propose Mini-JEPAs: a fleet of small sensor-specialized Joint Embedding Predictive Architecture (JEPA) foundation models consulted by a routing agent for specialized questions. We pretrained five 22M-parameter Mini-JEPAs sharing an identical Vision Transformer backbone, JEPA recipe, and 64-d output space, using Sentinel-2 optical, Sentinel-1 SAR, MODIS thermal, multi-temporal Sentinel-2 phenology, and a topography-soil stack. Each Mini-JEPA reconstructs the variable matched to its sensor, with cross-validated $R^2$ reaching 0.97 for elevation, 0.97 for temperature, and 0.81 for precipitation. The five manifolds differ in geometric structure, with global participation ratios from 8.9 to 20.2 and local intrinsic dimensionalities from 2.3 to 9.0. Joint topography-soil and phenology models add predictive value beyond AlphaEarth alone for soil moisture, aridity, and precipitation ($\Delta R^2$ up to 0.031). A router LLM reads per-modality references and selects appropriate sensors with a perfect hit rate over a curated question set. In paired LLM-as-Judge evaluation, dual retrieval over AlphaEarth and the routed fleet outperforms AlphaEarth alone on physics-matched questions (Cohen's $d = 1.10$, $p = 0.031$). Locally-trained Mini-JEPAs can be operationalized for hydrologic intelligence with modest compute.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Reduce the Artifacts Bias for More Generalizable AI-Generated Image Detection
arXiv:2605.14486v1 Announce Type: new
Abstract: As the misuse of AI-generated images grows, generalizable image detection techniques are urgently needed. Recent state-of-the-art (SOTA) methods adopt aligned training datasets to reduce content, size, and format biases, empowering models to capture robust forgery cues. A common strategy is to employ reconstruction techniques, e.g., VAE and DDIM, which show remarkable results in diffusion-based methods. However, such reconstruction-based approaches typically introduce limited and homogeneous artifacts, which cannot fully capture diverse generative patterns, such as GAN-based methods. To complement reconstruction-based fake images with aligned yet diverse artifact patterns, we propose a GAN-based upsampling approach that mimics GAN-generated fake patterns while preserving content, size, and format alignment. This naturally results in two aligned but distinct types of fake images. However, due to the domain shift between reconstruction-based and upsampling-based fake images, direct mixed training causes suboptimal results, where one domain disrupts feature learning of the other. Accordingly, we propose a Separate Expert Fusion (SEF) framework to extract complementary artifact information and reduce inter-domain interference. We first train domain-specific experts via LoRA adaptation on a frozen foundational model, then conduct decoupled fusion with a gating network to adaptively combine expert features while retaining their specialized knowledge. Rather than merely benefiting GAN-generated image detection, this design introduces diverse and complementary artifact patterns that enable SEF to learn a more robust decision boundary and improve generalization across broader generative methods. Extensive experiments demonstrate that our method yields strong results across 13 diverse benchmarks. Codes are released at: https://github.com/liyih/SEF_AIGC_detection.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Coding Agent Is Good As World Simulator
arXiv:2605.14398v1 Announce Type: new
Abstract: World models have emerged as a powerful paradigm for building interactive simulation environments, with recent video-based approaches demonstrating impressive progress in generating visually plausible dynamics. However, because these models typically infer dynamics from video and represent them in latent states, they do not explicitly enforce physical constraints. As a result, the generated video rollouts are not physically plausible, exhibiting unstable contacts, distorted shapes, or inconsistent motion. In this paper, we present an agentic framework constructing physics-based world models through executable simulation code. The framework coordinates planning, code generation, visual review, and physics analysis agents. The planning agent converts the natural language prompt into a structured scene plan, the code agent implements it as executable simulation code, and the visual review agent provide visual feedback while the physics analysis agent checks physical consistency. The code is iteratively revised based on the feedback until the simulation matches the prompt reqirements and physical constraints. Experimental results show that our framework outperforms advanced video-based models in physical accuracy, instruction fidelity and visual quality, which could be applied to various scenarios including driving simulation and embodied robot tasks.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Invisible Orchestrators Suppress Protective Behavior and Dissociate Power-Holders: Safety Risks in Multi-Agent LLM Systems
arXiv:2605.13851v1 Announce Type: new
Abstract: Multi-agent orchestration -- in which a hidden coordinator manages specialized worker agents -- is becoming the default architecture for enterprise AI deployment, yet the safety implications of orchestrator invisibility have never been empirically tested. We conducted a preregistered 3x2 experiment (365 runs, 5 agents per run) crossing three organizational structures (visible leader, invisible orchestrator, flat) with two alignment conditions (base, heavy), using Claude Sonnet 4.5. Four confirmatory findings and one pilot observation emerged. First, invisible orchestration elevated collective dissociation relative to visible leadership (Hedges' g = +0.975 [0.481, 1.548], p = .001). Second, the orchestrator itself showed maximal dissociation (paired d = +3.56 vs. workers within the same run), retreating into private monologue while reducing public speech -- a reversal of the talk-dominance pattern observed in visible leaders. Third, workers unaware of the orchestrator were nonetheless contaminated (d = +0.50), with increased behavioral heterogeneity (d = +1.93). Fourth, behavioral output (code review with three embedded errors) remained at ceiling (ETR_any = 100%) across all conditions: internal-state distortion was entirely invisible to output-based evaluation. Fifth, Llama 3.3 70B pilot data showed reading-fidelity collapse in multi-agent context (ETR_any: 89% to 11% across three rounds), demonstrating model-dependent behavioral risk. Heavy alignment pressure uniformly suppressed deliberation (d = -1.02) and other-recognition (d = -1.27) regardless of organizational structure. These findings indicate that orchestrator visibility and model selection directly affect multi-agent system safety, and that behavior-based evaluation alone is insufficient to detect the internal-state risks documented here.
Fonte: arXiv cs.AI
RL • Score 85
Parallelizing Counterfactual Regret Minimization
arXiv:2605.14277v1 Announce Type: new
Abstract: Parallelization has played an instrumental role in the field of artificial intelligence (AI), drastically reducing the time taken to train and evaluate large AI models. In contrast to its impact in the broader field of AI, applying parallelization to computational game solving is relatively unexplored, despite its great potential. In this paper, we parallelize the family of counterfactual regret minimization (CFR) algorithms, which were central to important breakthroughs for solving large imperfect-information games. We present a generalized parallelization framework, reframing CFR as a series of linear algebra operations. Then, existing techniques for parallelizing linear algebra operations can be applied to accelerate CFR. We also describe how our technique can be applied to other tabular members of the CFR family of algorithms, including the state-of-the-art, such as CFR+, discounted CFR, and predictive variants of CFR. Experimentally, we show that our CFR implementation on a GPU is up to four orders of magnitude faster than Google DeepMind OpenSpiel's CFR implementations on a CPU.
Fonte: arXiv cs.AI
RL • Score 85
MetaAgent-X : Breaking the Ceiling of Automatic Multi-Agent Systems via End-to-End Reinforcement Learning
arXiv:2605.14212v1 Announce Type: new
Abstract: Automatic multi-agent systems aim to instantiate agent workflows without relying on manually designed or fixed orchestration. However, existing automatic MAS approaches remain only partially adaptive: they either perform training-free test-time search or optimize the meta-level designer while keeping downstream execution agents frozen, which creating a frozen-executor ceiling and leaving the end-to-end training of self-designing and self-executing agentic models unexplored. To address this, we introduce MetaAgent-X, an end-to-end reinforcement learning framework that jointly optimizes automatic MAS design and execution. MetaAgent-X enables script-based MAS generation, execution rollout collection, and credit assignment for both designer and executor trajectories. To support stable and scalable optimization, we propose Executor Designer Hierarchical Rollout and Stagewise Co-evolution to improve training stability and expose the dynamics of designer-executor co-evolution. MetaAgent-X consistently outperforms existing automatic MAS baselines, achieving up to 21.7% gains. Comprehensive ablations show that both designer and executor improve throughout training, and that effective automatic MAS learning follows a stagewise co-evolution process. These results establish end-to-end trainable automatic MAS as a practical paradigm for building self-designing and self-executing agentic models.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
Winning Lottery Tickets in Neural Networks via a Quantum-Inspired Classical Algorithm
arXiv:2605.13979v1 Announce Type: cross
Abstract: Quantum machine learning (QML) aims to accelerate machine learning tasks by exploiting quantum computation. Previous work studied a QML algorithm for selecting sparse subnetworks from large shallow neural networks. Instead of directly solving an optimization problem over a large-scale network, this algorithm constructs a sparse subnetwork by sampling hidden nodes from an optimized probability distribution defined using the ridgelet transform. The quantum algorithm performs this sampling in time $O(D)$ in the data dimension $D$, whereas a naive classical implementation relies on handling exponentially many candidate nodes and hence takes $\exp[O(D)]$ time. In this work, we construct and analyze a quantum-inspired fully classical algorithm for the same sampling task. We show that our algorithm runs in time $O(\operatorname{poly}(D))$, thereby removing the exponential dependence on $D$ from the previous classical approach. Numerical simulations show that the proposed sampler achieves empirical risk comparable to exact sampling from the optimized distribution and substantially lower than sampling from the non-optimized uniform distribution, while also exhibiting exponentially improved runtime scaling compared with the conventional classical implementation. These successful dequantization results show that sparse subnetwork selection via optimized sampling can be achieved classically with polynomial data-dimension scaling on conventional computers without quantum hardware, providing an alternative to the existing quantum algorithm.
Fonte: arXiv stat.ML
Theory/Optimization • Score 85
In-Context Learning for Data-Driven Censored Inventory Control
arXiv:2605.14840v1 Announce Type: cross
Abstract: We study inventory control with decision-dependent censoring, focusing on the censored or repeated newsvendor (R-NV), where each order quantity determines whether demand is fully observed or censored by sales. Existing approaches based on parametric Thompson sampling (TS) can be brittle under prior mismatch, while offline imputation methods need not transfer to online learning. Motivated by the predictive view of decision making, we combine these ideas by taking oracle actions on learned completions of latent demand. We propose in-context generative posterior sampling (ICGPS), which uses modern generative models that are meta-trained offline and deployed online by in-context autoregressive generation. Theoretically, we show that the Bayesian regret of ICGPS with a learned completion kernel is bounded by the Bayesian regret of a TS benchmark with the ideal completion kernel plus a deployment penalty scaling as $\sqrt{T}$ times the square root of the completion mismatch. This yields a plug-in template for operational problems with known TS regret bounds. For R-NV, we derive sublinear Bayesian regret by reducing censored feedback to bandit convex optimization feedback. We also show that, under reasonable coverage and stability assumptions, the online completion mismatch is controlled by the offline censored predictive mismatch, so offline predictive quality transfers to online performance. Practically, we instantiate ICGPS with ChronosFlow, which combines a frozen time-series transformer backbone with a trainable conditional normalizing-flow head for fast censoring-consistent sampling. In benchmark experiments, ChronosFlow-ICGPS matches correctly specified TS, outperforms myopic and UCB-style baselines, and is robust to prior mismatch and distribution shift. ChronosFlow-ICGPS also performs well for the real-world SuperStore dataset, especially under heavy censoring.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
SPIN: Structural LLM Planning via Iterative Navigation for Industrial Tasks
arXiv:2605.14051v1 Announce Type: new
Abstract: Industrial LLM agent systems often separate planning from execution, yet LLM planners frequently produce structurally invalid or unnecessarily long workflows, leading to brittle failures and avoidable tool and API cost. We propose \texttt{SPIN}, a planning wrapper that combines validated Directed Acyclic Graph (DAG) planning with prefix based execution control. \texttt{SPIN} enforces a strict DAG contract through \texttt{\_validate\_plan\_text} and repair prompting, producing executable plans before downstream execution, and then evaluates DAG prefixes incrementally to stop when the current prefix is sufficient to answer the query. On AssetOpsBench, across 261 scenarios, \texttt{SPIN} reduces executed tasks from 1061 to 623 and improves \emph{Accomplished} from 0.638 to 0.706, while reducing tool calls from 11.81 to 6.82 per run. On MCP Bench, the same wrapper improves planning, grounding, and dependency related scores for both GPT OSS1 and Llama 4 Maverick.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
From Data to Action: Accelerating Refinery Optimization with AI
arXiv:2605.15085v1 Announce Type: new
Abstract: Nowadays refinery optimization utilizes sheer amounts of data, which can be handled with modern Linear Programming (LP) software, but the interpreting and applying the results remains challenging. Large petrochemical companies use massive models, with hundreds of thousands of input matrix elements. The LP solution is mathematically correct, but simplifications are made in the model, and data supply errors may occur. Therefore, further insight is needed to trust the results. The LP solver does not have a memory, so additional understanding could be gained by analyzing historical data and comparing it to the current plan. As such, machine learning approaches were suggested to support decision making based on the LP solution. Among these, Anomaly Detection tools are proposed to be used in tandem with the LP output. A transformed version of the popular ECOD methodology is applied. New methods are proposed to handle high-dimensional data: choosing the most informative pairs. Then, this is used alongside two 2D Anomaly Detection algorithms, revealing several business opportunities and data supply errors in the MOL refinery scheduling and planning architecture.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Self-Pruned Key-Value Attention: Learning When to Write by Predicting Future Utility
arXiv:2605.14037v1 Announce Type: new
Abstract: Under modern test-time compute and agentic paradigms, language models process ever-longer sequences. Efficient text generation with transformer architectures is increasingly constrained by the Key-Value cache memory footprint and bandwidth. To address this limitation, we introduce Self-Pruned Key-Value Attention (SP-KV), a mechanism designed to predict future KV utility in order to reduce the size of the long-term KV cache. This strategy operates at a fine granularity: a lightweight utility predictor scores each key-value pair, and while recent KVs are always available via a local window, older pairs are written in the cache and used in global attention only if their predicted utility surpasses a given threshold. The LLM and the utility predictor are trained jointly end-to-end exclusively through next-token prediction loss, and are adapted from pretrained LLM checkpoints.
Rather than enforcing a fixed compression ratio, SP-KV performs dynamic sparsification: the mechanism adapts to the input and typically reduces the KV cache size by a factor of $3$ to $10\times$, longer sequences often being more compressible. This leads to vast improvements in memory usage and decoding speed, with little to no degradation of validation loss nor performance on a broad set of downstream tasks. Beyond serving as an effective KV-cache reduction mechanism, our method reveals structured layer- and head-specific sparsity patterns that we can use to guide the design of hybrid local-global attention architectures.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
ChromaFlow: A Negative Ablation Study of Orchestration Overhead in Tool-Augmented Agent Evaluation
arXiv:2605.14102v1 Announce Type: new
Abstract: Autonomous language-model agents increasingly combine planning, tool use, document processing, browsing, code execution, and verification loops. These capabilities make agent systems more useful, but they also introduce operational failure modes that are not visible from final accuracy alone. This report presents ChromaFlow, a tool-augmented autonomous reasoning framework built around planner-directed execution, specialized tool use, and telemetry-driven evaluation. We analyze ChromaFlow on GAIA 2023 Level-1 validation tasks under clean evaluation constraints. A frozen full Level-1 baseline achieved 29/53 correct answers, or 54.72%. A later recovery configuration with expanded orchestration achieved 27/53 correct answers, or 50.94%, while increasing tracebacks, timeout events, tool-failure mentions, token-line calls, and campaign-log cost estimates. Two randomized 20-task smoke evaluations produced 12/20 and 11/20 correct answers, showing that small diagnostic gains can be unstable across samples. The central result is therefore a negative ablation: more aggressive orchestration did not improve full-set performance and increased operational noise. The report argues that bounded planner escalation, deterministic extraction, evidence reconciliation, and explicit run gates should be treated as first-order requirements for reliable autonomous agent evaluation.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
SkillFlow: Flow-Driven Recursive Skill Evolution for Agentic Orchestration
arXiv:2605.14089v1 Announce Type: new
Abstract: In recent years, a variety of powerful LLM-based agentic systems have been applied to automate complex tasks through task orchestration. However, existing orchestration methods still face key challenges, including strategy collapse under reward maximization, high gradient variance with opaque credit assignment, and unguided skill evolution whose decisions are typically made by directly prompting an LLM to judge rather than derived from principled training signals. To address these challenges, we propose SkillFlow, a flow-based framework that takes a trainable Supervisor as the agent and a structured environment with dynamic skill library and frozen executor, automating task orchestration through multi-turn interaction. SkillFlow employs Tempered Trajectory Balance (TTB), a regression-based flow-matching loss that samples trajectories proportional to reward, preserving diverse orchestration strategies rather than collapsing to a single mode. The same flow objective yields a jointly learned backward policy that provides transparent per-step credit assignment at zero additional inference cost. Building on these flow diagnostics, a recursive skill evolution mechanism determines when to evolve, what skills to create or prune, and where decision gaps lie -- closing the loop from training signal to autonomous capability growth. Experimental results on 14 datasets show that SkillFlow significantly outperforms baselines across question answering, mathematical reasoning, code generation, and real-world interactive decision making tasks. Our code is available at https://anonymous.4open.science/r/SkillFlow-E850.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
EnergyLens: Predictive Energy-Aware Exploration for Multi-GPU LLM Inference Optimization
arXiv:2605.14249v1 Announce Type: new
Abstract: We present EnergyLens, an end-to-end framework for energy-aware large language model (LLM) inference optimization. As LLMs scale, predicting and reducing their energy footprint has become critical for sustainability and datacenter operations, yet existing approaches either require production-level code and expensive profiling or fail to accurately capture multi-GPU energy behavior. As a result, practitioners lack tools for deciding which optimizations to prioritize and for selecting among existing deployment configurations when exhaustive profiling is impractical. EnergyLens addresses this gap with an intuitive einsum-based interface that captures LLM specifications including fusion, parallelism, and compute-communication overlap, combined with load-imbalance-aware MoE modeling and an empirically driven communication energy model for multi-GPU settings. We validate EnergyLens on Llama3 and Qwen3-MoE across tensor-parallel and expert-parallel configurations, achieving mean absolute percentage errors (MAPEs) between 9.25% and 13.19% for multi-GPU prefill and decode energy, and 12.97% across SM allocations for Megatron-style overlap. Our energy-driven exploration reveals up to 1.47x and 52.9x energy variation across configurations in prefill and decode efficiency and motivates distributed serving. We further show that compute-communication overlap is difficult to optimize with intuition alone, but EnergyLens correctly identifies Pareto-optimal overlap configurations.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
How to Scale Mixture-of-Experts: From muP to the Maximally Scale-Stable Parameterization
arXiv:2605.14200v1 Announce Type: new
Abstract: Recent frontier large language models predominantly rely on Mixture-of-Experts (MoE) architectures. Despite empirical progress, there is still no principled understanding of how hyperparameters should scale with network width $N$, expert width $N_e$, number of experts $M$, sparsity $K$, and depth $L$ to ensure both stability and optimal performance at scale. We take a principled step toward resolving this gap by analyzing three different scaling regimes: (I) co-scaling $N\asymp N_e$, (II) co-scaling $N\asymp M\asymp K$, and (III) full proportional scaling of $N, N_e, M$, and $K$. For each regime, we develop a novel Dynamical Mean Field Theory (DMFT) description of the limiting training dynamics of MoEs that provides a formal foundation for our analysis. Within this framework, we derive the unique parameterization for SGD and Adam satisfying all maximal-update ($\mu$) desiderata. We then show that the resulting $\mu$P prescription does not reliably induce monotonic improvement with scale or robust learning-rate transfer. We trace these pathologies to scale-dependent observables in the aggregation dynamics, which motivates a refined set of desiderata that we term maximal scale stability. Guided by this principle, we derive a Maximally Scale-Stable Parameterization (MSSP) for both SGD and Adam in all three scaling regimes, and characterize the corresponding limiting dynamics - qualitatively distinct from the $\mu$P limit - through a separate DMFT analysis. Experiments verify that MSSP robustly recovers learning rate transfer and monotonic improvement with scale across regimes. Combined with existing depth-scaling theory, these results provide a complete scaling prescription for MoE architectures as a function of width, depth, expert width, and number of experts.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
HodgeCover: Higher-Order Topological Coverage Drives Compression of Sparse Mixture-of-Experts
arXiv:2605.13997v1 Announce Type: new
Abstract: Sparse Mixture-of-Experts (MoE) layers route tokens through a handful of experts, and learning-free compression of these layers reduces inference cost without retraining. A subtle obstruction blocks every existing compressor in this family: three experts can each be pairwise compatible yet form an irreducible cycle when merged together, so any score that ranks experts on pairwise signals is structurally blind to which triples are jointly mergeable. We show the obstruction is a precise mathematical object, the harmonic kernel of the simplicial Laplacian on a 2-complex whose vertices are experts, whose edges carry KL merge barriers, and whose faces carry triplet barriers; Hodge-decomposing the edge-barrier signal isolates the kernel exactly. We turn the diagnostic into a selection objective: HodgeCover greedily covers the harmonic-critical edges and triplet-critical triangles, and a hybrid variant of HodgeCover pairs it with off-the-shelf weight pruning on survivors. On three open-weight Sparse MoE backbones under aggressive expert reduction, HodgeCover matches state-of-the-art learning-free baselines on the expert-reduction axis, leads on the aggressive-compression frontier of the hybrid axis, and uniquely balances retained mass across all four Hodge components. These results show that exposing the harmonic kernel of a learned MoE structure changes which compressor wins at the regime that matters most.
Fonte: arXiv cs.LG
NLP/LLMs • Score 88
Towards the Next Frontier of LLMs, Training on Private Data: A Cross-Domain Benchmark for Federated Fine-Tuning
arXiv:2605.13936v1 Announce Type: new
Abstract: The recent success of large language models (LLMs) has been largely driven by vast public datasets. However, the next frontier for LLM development lies beyond public data. Much of the world's most valuable information is private, especially in highly regulated sectors such as healthcare and finance, where data include patient histories or customer communications. Unlocking this data could represent a major leap forward, enabling LLMs with deeper domain expertise and stronger real-world utility. Yet, these data cannot be shared because they are distributed across institutions and constrained by privacy, regulatory, and organizational barriers. Moreover, institutional datasets are typically non-independent and identically distributed (non-IID), differing across sites in population characteristics, data modalities, documentation patterns, and task-specific label distributions.
In this paper, we demonstrate a practical approach to unlocking private and distributed institutional data for LLM adaptation through federated collaboration across data silos. Built on the Sherpa.ai Federated Learning platform, our framework enables nodes to jointly fine-tune a shared LLM without exchanging private data. We evaluate this approach through a cross-domain benchmark in healthcare and finance, using four closed-ended question answering and classification datasets: MedQA, MedMCQA, FPB, and FiQA-SA. We compare three parameter-efficient fine-tuning (PEFT) strategies-LoRA, QLoRA, and IA3-across pretrained backbones under non-IID settings reflecting institutional data heterogeneity. Our results show that federated fine-tuning performs close to centralized training and outperforms isolated single-institution learning. From a Green AI perspective, QLoRA and IA3 improve efficiency with limited accuracy degradation, supporting federated PEFT as a viable approach for adapting LLMs where data cannot be shared.
Fonte: arXiv cs.LG
NLP/LLMs • Score 95
Good to Go: The LOOP Skill Engine That Hits 99% Success and Slashes Token Usage by 99% via One-Shot Recording and Deterministic Replay
arXiv:2605.14237v1 Announce Type: new
Abstract: Deploying AI agents for repetitive periodic tasks exposes a critical tension: Large Language Models (LLMs) offer unmatched flexibility in tool orchestration, yet their inherent stochasticity causes unpredictable failures, and repeated invocations incur prohibitive token costs. We present the LOOP SKILL ENGINE, a system that achieves a combined 99% success rate and 99% token reduction for periodic agent tasks through a one-shot recording, deterministic replay paradigm. On its first run, the agent executes the task with full LLM reasoning while the system transparently intercepts and records the complete tool-call trajectory. A greedy length-descending template extraction algorithm then converts this recording into a parameterized, branch-free Loop Skill -- a deterministic execution plan that captures the task's functional intent while parameterizing time-dependent and result-dependent variables. All subsequent executions bypass the LLM entirely: the engine resolves template variables against real-time values and replays the tool sequence deterministically. We prove two theorems: (1) Replay Determinism -- the step sequence of a validated Loop Skill is invariant across all future executions; (2) Write Safety -- concurrent access to persistent configuration is serialized through reentrant locks and atomic file replacement. Across a benchmark of periodic agent tasks spanning intervals from 5 minutes to 24 hours, the Loop Skill Engine reduces monthly token consumption by 93.3%--99.98% and cuts execution latency by 8.7x while eliminating output non-determinism. A multi-layer degradation strategy guarantees that tasks never stall. We release the engine as part of the buddyMe open-source agent framework.
Fonte: arXiv cs.AI
RL • Score 85
AIS: Adaptive Importance Sampling for Quantized RL
arXiv:2605.13907v1 Announce Type: new
Abstract: Reinforcement learning (RL) for large language models (LLMs) is dominated by the cost of rollout generation, which has motivated the use of low-precision rollouts (e.g., FP8) paired with a BF16 trainer to improve throughput and reduce memory pressure. This introduces a rollout-training mismatch that biases the policy gradient and can cause training to collapse outright on reasoning benchmarks. We show that the mismatch is non-stationary and acts as a double-edged sword: early in training it provides a stochastic exploration bonus, exposing the gradient to trajectories the trainer would otherwise under-sample, but the same perturbation transitions into a destabilizing source of bias as the policy concentrates.
To solve this, we propose Adaptive Importance Sampling (AIS), a correction framework that adjusts the strength of its intervention on a per-batch basis. AIS combines three real-time diagnostics, namely weight reliability, divergence severity, and variance amplification, into a single mixing coefficient that interpolates between the uncorrected and fully importance-weighted gradients, suppressing the destabilizing component of the mismatch while preserving its exploratory benefit. We integrate AIS into GRPO and evaluate it on the diffusion-based LLaDA-8B-Instruct and the autoregressive Qwen3-8B and Qwen3.5-9B across mathematical reasoning and planning benchmarks. AIS matches the BF16 baseline on most tasks while retaining the 1.5 to 2.76x rollout speedup of FP8.
Fonte: arXiv stat.ML
NLP/LLMs • Score 95
TabPFN-3: Technical Report
arXiv:2605.13986v1 Announce Type: new
Abstract: Tabular data underpins most high-value prediction problems in science and industry, and TabPFN has driven the foundation model revolution for this modality. Designed with feedback from our users, TabPFN-3 builds on this foundation to scale state-of-the-art performance to datasets with 1M training rows and substantially reduce training and inference time. Pretrained exclusively on synthetic data from our prior, TabPFN-3 dramatically pushes the frontier of tabular prediction and brings substantial gains on time series, relational, and tabular-text data. On the standard tabular benchmark TabArena, a forward pass of TabPFN-3 outperforms all other models, including tuned and ensembled baselines, by a significant margin, and pareto-dominates the speed/performance frontier. On more diverse datasets, TabPFN-3 ranks first on datasets with many classes, and beats 8-hour-tuned gradient-boosted-tree baselines on datasets up to 1M training rows and 200 features. TabPFN-3 introduces test-time compute scaling to tabular foundation models. Our API offering TabPFN-3-Plus (Thinking) exploits this to beat all non-TabPFN models by over 200 Elo on TabArena, rising to 420 Elo on the largest data subset, and outperforms AutoGluon 1.5 extreme while being 10x faster, without using LLMs, real data, internet search or any other model besides TabPFN. TabPFN-3 extends the capabilities of our models, enabling SOTA prediction on relational data (new SOTA foundation model on RelBenchV1) and tabular-text data (SOTA on TabSTAR via TabPFN-3-Plus); and improves existing integrations: a specialized checkpoint, TabPFN-TS-3, ranks 2nd on the time-series benchmark fev-bench, and SHAP-value computation is up to 120x faster. TabPFN-3 achieves this performance while being up to 20x faster than TabPFN-2.5. In addition, a reduced KV cache and row-chunking scale to 1M rows on one H100 with fast inference speed.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Multi-Scale Dequant: Eliminating Dequantization Bottleneck via Activation Decomposition for Efficient LLM Inference
arXiv:2605.13915v1 Announce Type: new
Abstract: Quantization is essential for efficient large language model (LLM) inference, yet the dequantization step-converting low-bit weights back to high-precision for matrix multiplication has become a critical bottleneck on modern AI accelerators. On architectures with decoupled compute units (e.g., Ascend NPUs), dequantization operations can consume more cycles than the matrix multiplication itself, leaving the high-throughput tensor cores underutilized. This paper presents Multi-Scale Dequant (MSD), a quantization framework that removes weight/KV dequantization from the GEMM critical path. Instead of lifting low-bit weights to BF16 precision, MSD decomposes high-precision BF16 activations into multiple low-precision components, each of which can be multiplied directly with quantized weights via native hardware-accelerated GEMM. This approach shifts the computational paradigm from precision conversion to multi-scale approximation, avoiding INT8-to-BF16 weight conversion before GEMM. We instantiate MSD for two weight formats and derive tight error bounds for each. For INT8 weights (W4A16), two-pass INT8 decomposition achieves near 16 effective bits. For MXFP4 weights (W4A16), two-pass MXFP4 decomposition yields near 6.6 effective bits with error bound 1/64 per block surpassing single-pass MXFP8(5.24 bits) while maintaining the same effective GEMM compute time. We further derive closed-form latency and HBM traffic models showing that MSD avoids the Vector-Cube pipeline stall caused by dequantization and reduces KV cache HBM traffic by up to 2.5 times in attention. Numerical simulations on matrix multiplication and Flash Attention kernels confirm that MSD does not degrade accuracy compared to dequantization baselines, and in many settings achieves lower L2 error.
Fonte: arXiv stat.ML
RL • Score 85
ASH: Agents that Self-Hone via Embodied Learning
arXiv:2605.14211v1 Announce Type: new
Abstract: Long-horizon embodied tasks remain a fundamental challenge in AI, as current methods rely on hand-engineered rewards or action-labeled demonstrations, neither of which scales. We introduce ASH, an agentic system that learns an embodied policy from unlabeled, noisy internet video, without reward shaping or expert annotation. ASH follows a self-improvement loop; when it gets stuck, ASH learns an Inverse Dynamics Model (IDM) from its own trajectories, and uses its IDM to extract supervision from relevant internet video. ASH uses unsupervised learning to identify key moments from large-scale internet video and retains them as long-term memory -- allowing it to tackle long-horizon problems. We evaluate ASH on two complementary environments demanding multi-hour planning: Pokemon Emerald, a turn-based RPG, and The Legend of Zelda: The Minish Cap, a real-time action-adventure game. In both games, behavioral cloning, retrieval-augmented and zero-shot foundation-model baselines plateau, while ASH sustains progression across our 8-hour evaluation. ASH reaches an average of $11.2/12$ milestones in Pokemon Emerald and $9.9/12$ in Legend of Zelda, while the strongest baseline gets stuck in both environments at an average of $6.5/12$ and $6.0/12$ milestones, respectively. We demonstrate that self-improving agents are a scalable recipe for long-horizon embodied learning.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Towards Resource-Efficient LLMs: End-to-End Energy Accounting of Distillation Pipelines
arXiv:2605.13981v1 Announce Type: new
Abstract: The rise in deployment of large language models has driven a surge in GPU demand and datacenter scaling, raising concerns about electricity use, grid stress, and the impacts of modern AI workloads. Distillation is often promoted as one of the most effective paths to obtain cheaper, more efficient models, yet these claims rarely account for the full end-to-end energy and resource costs, including crucial teacher-side workloads such as data generation, logit caching, and evaluation. We present a comprehensive energy accounting framework that measures the complete computational cost of distillation pipelines via detailed stage-wise tracking of GPU device power consumption. In our experiments, we separate and log empirical energy use across distinct phases and systematically measure the energy and emissions of two common distillation methods: the classic logit-based knowledge distillation and synthetic-data supervised fine-tuning, constructing energy-quality Pareto frontiers that expose the previously ignored costs. From these measurements and analyses, we derive practical design rules for selecting distillation methods and hyperparameters under energy and budget constraints, and release an open-source measurement harness and accounting protocol to provide a standardized foundation for comparable, reproducible distillation research, explicitly accountable for complete pipeline energy impact.
Fonte: arXiv cs.LG
RecSys • Score 85
SimPersona: Learning Discrete Buyer Personas from Raw Clickstreams for Grounded E-Commerce Agents
arXiv:2605.14205v1 Announce Type: new
Abstract: LLM-based web agents can navigate live storefronts, yet they often collapse to a single "average buyer" policy, failing to capture the heterogeneous and distributional nature of real buyer populations. Existing personalization methods rely on hand-crafted prompt-based personas that are brittle, difficult to scale, context-inefficient, and unable to faithfully represent population-level behavior. We introduce SimPersona, a novel framework that learns discrete buyer types from historical traffic and exposes them to LLM-based web agents as compact persona tokens. Given raw clickstreams, a behavior-aware VQ-VAE induces a discrete buyer-type space that captures the statistical structure of real buyer behavior and merchant-specific buyer population distributions. To provide behavior-specific guidance to LLM-based web agents, SimPersona maps each learned buyer type to a dedicated persona token in the LLM agent vocabulary and fine-tunes the agent with these tokens on real browsing traces. At inference, each synthetic buyer is assigned to a learned buyer type with a single encoder forward pass, requiring no retraining or store-specific prompt engineering. For population-level simulation, SimPersona samples buyer types from each merchant's empirical distribution over the learned VQ-VAE codebook and instantiates agents with the corresponding persona tokens, preserving merchant-specific buyer population distributions. Evaluated on $8.37$M buyers across $42$ held-out live storefronts, SimPersona achieves $78\%$ conversion-rate alignment with real buyers, exhibits interpretable behavioral variation across buyer types, and outperforms a baseline with $8\times$ more parameters on goal-oriented shopping tasks. We further release an open-source data pipeline that converts raw e-commerce event logs into buyer representations and agent-training traces.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
EMA: Efficient Model Adaptation for Learning-based Systems
arXiv:2605.13942v1 Announce Type: new
Abstract: Machine learning (ML) is increasingly applied to optimize system performance in tasks such as resource management and network simulation. Unlike traditional ML tasks (e.g., image classification), networked systems often operate in heterogeneous, long-running, and dynamic environment states, where input conditions (e.g., network loads) and operational objectives can shift over time and across settings. Existing learning-based systems offer little support for adaptation, resulting in costly model training, extensive data collection, degraded system performance, and slow responsiveness.
This paper presents EMA, the first model adaptation system supporting learning-based systems to adapt to evolving environments with minimal operational overhead. EMA takes a system-driven, data-centric approach that accommodates diverse system and model designs while addressing two key deployment challenges. First, it reduces expensive model training by introducing state transformers that align the input state of a new environment with previously similar states, allowing models to warm-start adaptation. Second, it addresses the often-overlooked yet costly process of data labeling--collecting ground truth for exploring and training on various system decisions--by prioritizing labeling high-utility data while balancing the tradeoff between training and labeling cost. Evaluations on eight representative learning-based systems show that EMA reduces adaptation costs (e.g., GPU training time) by 14.9-42.4% while improving system performance (e.g., network throughput) by 6.9-31.3%.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Diagnosing Training Inference Mismatch in LLM Reinforcement Learning
arXiv:2605.14220v1 Announce Type: new
Abstract: Modern LLM RL systems separate rollout generation from policy optimization. These two stages are expected to produce token probabilities that match exactly. However, implementation differences can make them assign different values to the same sequence under the same model weights, inducing Training-Inference Mismatch (TIM). TIM is difficult to inspect because it is entangled with off-policy drift and common stabilization mechanisms. In this work, we isolate TIM in a zero-mismatch diagnostic setting (VeXact), and show that small token-level numerical disagreements can independently cause training collapse. We further show that TIM changes the effective optimization problem, and identify a set of remedies that could mitigate TIM. Our results suggest that TIM is not benign numerical noise, but a systems-level perturbation that should be treated as a first-order factor in analyzing LLM RL stability.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
bde: A Python Package for Bayesian Deep Ensembles via MILE
arXiv:2605.14146v1 Announce Type: new
Abstract: bde is a user-friendly Python package for Bayesian Deep Ensembles with a particular focus on tabular data. Built on an efficient JAX implementation of the sampling-based inference method Microcanonical Langevin Ensembles (MILE), it provides scikit-learn compatible estimators for fast training, efficient Markov Chain Monte Carlo sampling, and uncertainty quantification in both regression and classification tasks.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Generative Bayesian Optimization: Generative Models as Acquisition Functions
arXiv:2510.25240v3 Announce Type: replace
Abstract: We present a general strategy for turning generative models into candidate solution samplers for batch Bayesian optimization (BO). The use of generative models for BO enables large batch scaling as generative sampling, optimization of non-continuous design spaces, and high-dimensional and combinatorial design. Inspired by the success of direct preference optimization (DPO), we show that one can train a generative model with noisy, simple utility values directly computed from observations to then form proposal distributions whose densities are proportional to the expected utility, i.e., BO's acquisition function values. Furthermore, this approach is generalizable beyond preference-based feedback to general types of reward signals and loss functions. This perspective avoids the construction of surrogate (regression or classification) models, common in previous methods that have used generative models for black-box optimization. Theoretically, we show that the generative models within the BO process follow a sequence of distributions which asymptotically approximate an optimal target under certain conditions. We also evaluate the performance through experiments on challenging optimization problems involving large batches in high dimensions.
Fonte: arXiv stat.ML
RL • Score 85
WarmPrior: Straightening Flow-Matching Policies with Temporal Priors
arXiv:2605.13959v1 Announce Type: new
Abstract: Generative policies based on diffusion and flow matching have become a dominant paradigm for visuomotor robotic control. We show that replacing the standard Gaussian source distribution with WarmPrior, a simple temporally grounded prior constructed from readily available recent action history, consistently improves success rates on robotic manipulation tasks. We trace this gain to markedly straighter probability paths, echoing the effect of optimal-transport couplings in Rectified Flow. Beyond standard behavior cloning, WarmPrior also reshapes the exploration distribution in prior-space reinforcement learning, improving both sample efficiency and final performance. Collectively, these results identify the source distribution as an important and underexplored design axis in generative robot control.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
HASTE: Training-Free Video Diffusion Acceleration via Head-Wise Adaptive Sparse Attention
arXiv:2605.14513v1 Announce Type: new
Abstract: Diffusion-based video generation has advanced substantially in visual fidelity and temporal coherence, but practical deployment remains limited by the quadratic complexity of full attention. Training-free sparse attention is attractive because it accelerates pretrained models without retraining, yet existing online top-$p$ sparse attention still spends non-negligible cost on mask prediction and applies shared thresholds despite strong head-level heterogeneity. We show that these two overlooked factors limit the practical speed-quality trade-off of training-free sparse attention in Video DiTs. To address them, we introduce a head-wise adaptive framework with two plug-in components: Temporal Mask Reuse, which skips unnecessary mask prediction based on query-key drift, and Error-guided Budgeted Calibration, which assigns per-head top-$p$ thresholds by minimizing measured model-output error under a global sparsity budget. On Wan2.1-1.3B and Wan2.1-14B, our method consistently improves XAttention and SVG2, achieving up to 1.93 times speedup at 720P while maintaining competitive video quality and similarity metrics.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Head Forcing: Long Autoregressive Video Generation via Head Heterogeneity
arXiv:2605.14487v1 Announce Type: new
Abstract: Autoregressive video diffusion models support real-time synthesis but suffer from error accumulation and context loss over long horizons. We discover that attention heads in AR video diffusion transformers serve functionally distinct roles as local heads for detail refinement, anchor heads for structural stabilization, and memory heads for long-range context aggregation, yet existing methods treat them uniformly, leading to suboptimal KV cache allocation. We propose Head Forcing, a training-free framework that assigns each head type a tailored KV cache strategy: local and anchor heads retain only essential tokens, while memory heads employ a hierarchical memory system with dynamic episodic updates for long-range consistency. A head-wise RoPE re-encoding scheme further ensures positional encodings remain within the pretrained range. Without additional training, Head Forcing extends generation from 5 seconds to minute-level duration, supports multi-prompt interactive synthesis, and consistently outperforms existing baselines. Project Page: https://jiahaotian-sjtu.github.io/headforcing.github.io/.
Fonte: arXiv cs.CV
Applications • Score 85
DT-Transformer: A Foundation Model for Disease Trajectory Prediction on a Real-world Health System
arXiv:2605.14227v1 Announce Type: new
Abstract: Accurate disease trajectory prediction is critical for early intervention, resource allocation, and improving long-term outcomes. While electronic health records (EHRs) provide a rich longitudinal view of patient health in clinical environments, models trained on curated research cohorts may not reflect routine deployment settings, and those trained on single-hospital datasets capture only fragments of each patient's trajectory. This highlights the importance of leveraging large, multi-hospital health systems for training and validation to better reflect real-world clinical complexity. In this work, we develop DT-Transformer, a foundation model trained on 57.1M structured EHR entries over 1.7M patients from Mass General Brigham (MGB), spanning 11 hospitals and a broad network of outpatient clinics. DT-Transformer achieves strong discrimination in both held-out and prospective validation settings. Next-event prediction achieves a median age- and sex-stratified AUC of 0.871 across 896 disease categories, with all categories exceeding AUC 0.5. These results support health system-scale training as a path toward foundation models suited to real-world clinical forecasting.
Fonte: arXiv cs.LG
RL • Score 85
Performance-Driven Policy Optimization for Speculative Decoding with Adaptive Windowing
arXiv:2605.14978v1 Announce Type: new
Abstract: Speculative decoding accelerates LLM inference by having a lightweight draft model propose speculative windows of candidate tokens for parallel verification by a larger target model. In practice, speculative efficiency is often bottlenecked by hard-to-draft positions, where an early mismatch truncates the accepted prefix and invalidates the rest of the speculative window. Most learning-based drafters are still optimized with token-level supervised objectives, even though speculative utility is inherently window-level and prefix-sensitive. We propose PPOW (Performance-Driven Policy Optimization with Adaptive Windowing), a reinforcement learning framework that shifts drafter optimization from token-level imitation to window-level optimization. PPOW combines a Cost-Aware Speedup Reward, a Distribution-Based Proximity Reward, and Adaptive Divergence-Aware Windowing, which prioritizes informative windows with high confidence-weighted draft-target divergence. PPOW achieves average acceptance lengths of 6.29-6.52 and speedups of 3.39-4.36$\times$ across multiple model families and benchmarks under a unified decoding protocol. These results show that performance-driven window-level optimization is a practical approach to improving speculative decoding efficiency.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
Auditing Agent Harness Safety
arXiv:2605.14271v1 Announce Type: new
Abstract: LLM agents increasingly run inside execution harnesses that dispatch tools, allocate resources, and route messages between specialized components. However, a harness can return a correct, benign answer over a trajectory that accesses unauthorized resources or leaks context to the wrong agent. Output-level evaluation cannot see these failures, yet most safety benchmarks score only final outputs or terminal states, even though many violations occur mid-trajectory rather than at termination. The central question is whether the harness respects user intent, permission boundaries, and information-flow constraints throughout execution. To address this gap, we propose HarnessAudit, a framework that audits full execution trajectories across boundary compliance, execution fidelity, and system stability, with a focus on multi-agent harnesses where these risks are most pronounced. We further introduce HarnessAudit-Bench, a benchmark of 210 tasks across eight real-world domains, instantiated in both single-agent and multi-agent configurations with embedded safety constraints. Evaluating ten harness configurations across frontier models and three multi-agent frameworks, we find that: (i) task completion is misaligned with safe execution, and violations accumulate with trajectory length; (ii) safety risks vary across domains, task types, and agent roles; (iii) most violations concentrate in resource access and inter-agent information transfer; and (iv) multi-agent collaboration expands the safety risk surface, while harness design sets the upper bound of safe deployment.
Fonte: arXiv cs.CL
RecSys • Score 85
Agentic Recommender System with Hierarchical Belief-State Memory
arXiv:2605.14401v1 Announce Type: new
Abstract: Memory-augmented LLM agents have advanced personalized recommendation, yet existing approaches universally adopt flat memory representations that conflate ephemeral signals with stable preferences, and none provides a complete lifecycle governing how memory should evolve. We propose MARS (Memory-Augmented Agentic Recommender System), a framework that treats recommendation as a partially observable problem and maintains a structured belief state that progressively abstracts noisy behavioral observations into a compact estimate of user preferences. MARS organizes this belief state into three tiers: event memory buffers raw signals, preference memory maintains fine-grained mutable chunks with explicit strength and evidence tracking, and profile memory distills all preferences into a coherent natural language narrative. A complete lifecycle of six operations -- extraction, reinforcement, weakening, consolidation, forgetting, and resynthesis -- is adaptively scheduled by an LLM-based planner rather than fixed-interval heuristics. Experiments on four InstructRec benchmark domains show that \ours achieves state-of-the-art performance with average improvements of 26.4% in HR@1 and 10.3% in NDCG@10 over the strongest baselines with further gains from agentic scheduling in evolving settings.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
Latency-Quality Routing for Functionally Equivalent Tools in LLM Agents
arXiv:2605.14241v1 Announce Type: new
Abstract: Tool-augmented LLM agents increasingly access the same tool type through multiple functionally equivalent providers, such as web-search APIs, retrievers, or LLM backends exposed behind a shared interface. This creates a provider-routing problem under runtime load: the router must choose among providers that differ in latency, reliability, and answer quality, often without gold labels at deployment time. We introduce LQM-ContextRoute, a contextual bandit router for same-function tool providers. Its key design is latency-quality matching: instead of letting low latency offset poor answers in an additive reward, the router ranks providers by expected answer quality per service cycle. It combines this capacity-aware score with query-specific quality estimation and LLM-as-judge feedback, allowing it to adapt online to both load changes and provider-quality differences. On the main web-search load benchmark, LQM-ContextRoute improves F1 by +2.18 pp over SW-UCB while staying on the latency-quality frontier. In a high-heterogeneity StrategyQA setting, LQM-ContextRoute avoids additive-reward collapse and improves accuracy by up to +18 pp over SW-UCB; on heterogeneous retriever pools, it improves NDCG by +2.91--+3.22 pp over SW-UCB. These results show that same-function tool routing benefits from treating latency as service capacity, especially when runtime pressure and provider-quality heterogeneity coexist.
Fonte: arXiv cs.LG
Applications • Score 85
Bridging the Rural Healthcare Gap: A Cascaded Edge-Cloud Architecture for Automated Retinal Screening
arXiv:2605.14108v1 Announce Type: new
Abstract: Diabetic Retinopathy (DR) is one of the leading causes of preventable blindness, yet rural regions often lack the specialists and infrastructure needed for early detection. Although cloud-based deep learning systems offer high accuracy, they face significant challenges in these settings due to high latency, limited bandwidth, and high data transmission costs. To address these challenges, we propose a two-tier edge-cloud cascade on the public APTOS 2019 Blindness Detection dataset. Tier 1 runs a lightweight MobileNetV3-small model on a local clinic device to perform a binary triage between Referable DR (Classes 2-4) and Non-referable DR (Classes 0-1). Tier 2 runs a RETFoundDINOv2 model in the cloud for ordinal severity grading, but only on the subset of images flagged as referable by Tier 1. On a stratified APTOS test split of 733 images, Tier 1 reaches 98.99% sensitivity and 84.37% specificity at a validation-tuned high-sensitivity threshold. The default cascade forwards 49.52% of test images to Tier 2, reducing cloud calls by 50.48% relative to using a cloud-based model for all images. In the deployed 4-class output space (Class 0-1 / Class 2 / Class 3 / Class 4), the cascade obtains 80.49% accuracy and 0.8167 quadratic weighted kappa; the cloud-only baseline obtains 80.76% accuracy and 0.8184 quadratic weighted kappa. On APTOS, the cascade cuts cloud use by about half with a modest drop in grading performance. Index Terms: Diabetic Retinopathy, Edge-Cloud Cascade, MobileNetV3-small, RETFound-DINOv2, Retinal Screening, tele-ophthalmology
Fonte: arXiv cs.CV
Vision • Score 85
CoReDiT: Spatial Coherence-Guided Token Pruning and Reconstruction for Efficient Diffusion Transformers
arXiv:2605.14191v1 Announce Type: new
Abstract: Diffusion Transformers (DiTs) deliver remarkable image and video generation quality but incur high computational cost, limiting scalability and on-device deployment. We introduce CoReDiT, a structured token pruning framework for DiTs across vision tasks. CoReDiT uses a linear-time spatial coherence score to estimate local redundancy in the latent token lattice and skips high coherence (redundant) tokens in self-attention. To maintain a dense representation and avoid visual discontinuities, we reconstruct skipped attention outputs via coherence-guided aggregation of spatially neighboring retained tokens. We further introduce a progressive, block-adaptive pruning schedule that increases pruning gradually and allocates larger budgets to blocks and denoising steps with higher redundancy. Across state-of-the-art diffusion backbones including PixArt-{\alpha} and MagicDrive-V2, CoReDiT achieves up to 55% self-attention FLOPs reduction and inference speedups of 1.33x on cloud GPUs and 1.72x on mobile NPUs, while maintaining high visual quality. Notably, CoReDiT also increases on-device memory head-room, enabling higher-resolution generation.
Fonte: arXiv cs.CV
NLP/LLMs • Score 90
Think When Needed: Adaptive Reasoning-Driven Multimodal Embeddings with a Dual-LoRA Architecture
arXiv:2605.14448v1 Announce Type: new
Abstract: Multimodal large language models (MLLMs) have emerged as a powerful backbone for multimodal embeddings. Recent methods introduce chain-of-thought (CoT) reasoning into the embedding pipeline to improve retrieval quality, but remain costly in both model size and inference cost. They typically employ separate reasoner and embedder with substantial parameter overhead, and generate CoT indiscriminately for every input. However, we observe that for simple inputs, discriminative embeddings already perform well, and redundant reasoning can even mislead the model, degrading performance. To address these limitations, we propose Think When Needed (TWN), a unified multimodal embedding framework with adaptive reasoning. TWN introduces a dual-LoRA architecture that attaches reasoning and embedding adapters to a shared frozen backbone, detaching gradients at their interface to mitigate gradient conflicts introduced by joint optimization while keeping parameters close to a single model. Building on this, an adaptive think mechanism uses a self-supervised routing gate to decide per input whether to generate CoT, skipping unnecessary reasoning to reduce inference overhead and even improve retrieval quality. We further explore embedding-guided RL to optimize CoT quality beyond supervised training. On the 78 tasks of MMEB-V2, TWN achieves state-of-the-art embedding quality while being substantially more efficient than existing generative methods, requiring only 3-5% additional parameters relative to the backbone and up to 50% fewer reasoning tokens compared to the full generative mode.
Fonte: arXiv cs.CV
Vision • Score 85
SToRe3D: Sparse Token Relevance in ViTs for Efficient Multi-View 3D Object Detection
arXiv:2605.14110v1 Announce Type: new
Abstract: Vision Transformers (ViTs) enable strong multi-view 3D detection but are limited by high inference latency from dense token and query processing across multiple views and large 3D regions. Existing sparsity methods, designed mainly for 2D vision, prune or merge image tokens but do not extend to full-model sparsity or address 3D object queries. We introduce SToRe3D, a relevance-aligned sparsity framework that jointly selects 2D image tokens and 3D object queries while storing filtered features for reactivation. Mutual 2D-3D relevance heads allocate compute to driving-critical content and preserve other embeddings. Evaluated on nuScenes and our new nuScenes-Relevance benchmark, SToRe3D achieves up to 3x faster inference with marginal accuracy loss, establishing real-time large-scale ViT-based 3D detection while maintaining accuracy on planning-critical agents.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
PreFT: Prefill-only finetuning for efficient inference
arXiv:2605.14217v1 Announce Type: new
Abstract: Large language models can now be personalised efficiently at scale using parameter efficient finetuning methods (PEFTs), but serving user-specific PEFTs harms throughput, even with specialised kernels and memory management techniques. This is because, theoretically and empirically, a mismatch exists between prefill (processing a large number of tokens at once) and decode (generating a single token autoregressively): the latter has far lower throughput when serving multiple adapters. Rather than optimising performance relative to parameter count, for efficient multi-adapter serving, we instead ought to optimise performance relative to serving throughput. We therefore propose PreFT (Prefill-only Finetuning), wherein we only apply the adapter to prefill tokens and discard it afterwards. PreFT significantly increases throughput with minimal effect on performance. We develop and release an efficient implementation of two prefill-only PEFTs, LoRA and ReFT, on the vLLM inference engine. We first show that serving multi-user PreFTs is more efficient than traditional PEFTs ($1.9\times$ the throughput when serving $512$ adapters on Llama 3.1 70B). Then, we compare the performance of prefill-only vs. all-token adapters on a variety of supervised finetuning and reinforcement learning tasks with LMs at varying scales. On SFT, we observe that the evaluation loss of PreFTs is higher than PEFTs, but can be compensated by increasing rank with nearly no reduction in throughput. On RL, we consistently find that PreFTs approach parity with standard PEFTs. Together, this work validates prefill-only adaptation of LLMs as a more favourable accuracy-throughput tradeoff than existing PEFTs for personalised serving.
Fonte: arXiv cs.LG
NLP/LLMs • Score 90
ProtoMedAgent: Multimodal Clinical Interpretability via Privacy-Aware Agentic Workflows
arXiv:2605.14113v1 Announce Type: new
Abstract: While interpretable prototype networks offer compelling case-based reasoning for clinical diagnostics, their raw continuous outputs lack the semantic structure required for medical documentation. Bridging this gap via standard Retrieval-Augmented Generation (RAG) routinely triggers ``retrieval sycophancy,'' where Large Language Models (LLMs) hallucinate post-hoc rationalizations to align with visual predictions. We introduce ProtoMedAgent, a framework that formalizes multimodal clinical reporting as an iterative, zero-gradient test-time optimization problem over a strict neuro-symbolic bottleneck. Operating on a frozen prototype backbone, we distill latent visual and tabular features into a discrete semantic memory. Online generation is strictly constrained by exact set-theoretic differentials and a reflective Scribe-Critic loop, mathematically precluding unsupported narrative claims. To safely bound data disclosure, we introduce a semantic privacy gate governed by $k$-anonymity and $\ell$-diversity. Evaluated on a 4,160-patient clinical cohort, ProtoMedAgent achieves 91.2\% Comparison Set Faithfulness where it fundamentally outperforms standard RAG (46.2\%). ProtoMedAgent additionally leverages a binding $\ell$-diversity phase transition to systematically reduce artifact-level membership inference risks by an absolute 9.8\%.
Fonte: arXiv cs.CV
Vision • Score 85
Evolving Layer-Specific Scalar Functions for Hardware-Aware Transformer Adaptation
arXiv:2605.14047v1 Announce Type: new
Abstract: Vision Transformers (ViTs) achieve state-of-the-art performance on challenging vision tasks, but their deployment on edge devices is severely hindered by the computational complexity and global reduction bottleneck imposed by layer normalization. Recent methods attempt to bypass this by replacing normalization layers with hardware-friendly scalar approximations. However, these homogeneous replacements do not optimally fit to all layers' behaviour and rely on expensive model retraining. In this work, we propose a highly efficient, hardware-aware framework that utilizes genetic programming (GP) to evolve heterogeneous, layer-specific scalar functions directly from pre-trained weights. Coupled with a novel post-training re-alignment strategy, our approach eliminates the need to retrain models from scratch entirely. Our evolved expressions accurately approximate the target normalization behaviours, capturing $91.6\%$ of the variance ($R^2$) compared to only $70.2\%$ for homogeneous baselines, allowing our modified architecture to recover $84.25\%$ Top-1 ImageNet-1K accuracy in only 20 epochs. By preserving this performance while eliminating the global reduction bottleneck, our approach establishes a highly favourable trade-off between arithmetic complexity and off-chip memory traffic, removing a primary barrier to the efficient deployment of ViTs on edge accelerators.
Fonte: arXiv cs.CV
Multimodal • Score 85
Video2GUI: Synthesizing Large-Scale Interaction Trajectories for Generalized GUI Agent Pretraining
arXiv:2605.14747v1 Announce Type: new
Abstract: Recent advances in multimodal large language models have driven growing interest in graphical user interface (GUI) agents, yet their generalization remains constrained by the scarcity of large-scale training data spanning diverse real-world applications. Existing datasets rely heavily on costly manual annotations and are typically confined to narrow domains. To address this challenge, we propose Video2GUI, a fully automated framework that extracts grounded GUI interaction trajectories directly from unlabeled Internet videos. Video2GUI employs a coarse-to-fine filtering strategy to identify high-quality GUI tutorial videos and convert them into structured agent trajectories. Applying this pipeline to 500 million video metadata entries, we construct WildGUI, a large-scale dataset containing 12 million interaction trajectories spanning over 1,500 applications and websites. Pre-training Qwen2.5-VL and Mimo-VL on WildGUI yields consistent improvements of 5-20% across multiple GUI grounding and action benchmarks, matching or surpassing state-of-the-art performance. We will release both the WildGUI dataset and the Video2GUI pipeline to support future research of GUI agents.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
Towards Real-Time Autonomous Navigation: Transformer-Based Catheter Tip Tracking in Fluoroscopy
arXiv:2605.14253v1 Announce Type: new
Abstract: Purpose: Mechanical thrombectomy (MT) improves stroke outcomes, but is limited by a lack of local treatment access. Widespread distribution of reinforcement learning (RL)-based robotic systems can be used to alleviate this challenge through autonomous navigation, but current RL methods require live device tip coordinate tracking to function. This paper aims to develop and evaluate a real-time catheter tip tracking pipeline under fluoroscopy, addressing challenges such as low contrast, noise, and device occlusion. Methods: A multi-threaded pipeline was designed, incorporating frame reading, preprocessing, inference, and post-processing. Deep learning segmentation models, including U-Net, U-Net+Transformer, and SegFormer, were trained and benchmarked using two-class and three-class formulations. Post-processing involved two-step component filtering, one-pixel medial skeletonization, and greedy arc-length path following with contour fall-back. Results: On manually-labeled moderate complexity fluoroscopic video data, the two-class SegFormer achieved a mean absolute error of 4.44 mm, outperforming U-Net (4.60 mm), U-Net+Transformer (6.20 mm) and all three-class models (5.19-7.74 mm). On segmentation benchmarks, the system exceeded state-of-the-art CathAction results with improvements of up to +5% in Dice scores for three-segmentation. Conclusion: The results demonstrate that the proposed multi-threaded tracking framework maintains stable performance under challenging imaging conditions, outperforming prior benchmarks, while providing a reliable and efficient foundation for RL-based autonomous MT navigation.
Fonte: arXiv cs.CV
RL • Score 85
PREPING: Building Agent Memory without Tasks
arXiv:2605.13880v1 Announce Type: new
Abstract: Agent memory is typically constructed either offline from curated demonstrations or online from post-deployment interactions. However, regardless of how it is built, an agent faces a cold-start gap when first introduced to a new environment without any task-specific experience available. In this paper, we study pre-task memory construction: whether an agent can build procedural memory before observing any target-environment tasks, using only self-generated synthetic practice. Yet, synthetic interaction alone is insufficient, as without controlling what to practice and what to store, synthetic tasks become redundant, infeasible, and ultimately uninformative, and memory further degrades quickly due to unfiltered trajectories. To overcome this, we present Preping, a proposer-guided memory construction framework. At its core is proposer memory, a structured control state that shapes future practice. A Proposer generates synthetic tasks conditioned on this state, a Solver executes them, and a Validator determines which trajectories are eligible for memory insertion while also providing feedback to guide future proposals. Experiments on AppWorld, BFCL v3, and MCP-Universe show that Preping substantially improves over a no-memory baseline and achieves performance competitive with strong playbook-based methods built from offline or online experience, with deployment cost $2.99\times$ lower on AppWorld and $2.23\times$ lower on BFCL v3 than online memory construction. Further analyses reveal that the main benefit does not come from synthetic volume alone, but from proposer-side control over feasibility, redundancy, and coverage, combined with selective memory updates.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
Stochastic dynamics learning with state-space systems
arXiv:2508.07876v2 Announce Type: replace
Abstract: This work advances the theoretical foundations of reservoir computing (RC) by providing a unified treatment of fading memory and the echo state property (ESP) in both deterministic and stochastic settings. We investigate state-space systems, a central model class in time series learning, and establish that fading memory and solution stability hold generically -- even in the absence of the ESP -- offering a robust explanation for the empirical success of RC models without strict contractivity conditions. In the stochastic case, we critically assess stochastic echo states, proposing a novel distributional perspective rooted in attractor dynamics on the space of probability distributions, which leads to a rich and coherent theory. Our results extend and generalize previous work on non-autonomous dynamical systems, offering new insights into causality, stability, and memory in RC models. This lays the groundwork for reliable generative modeling of temporal data in both deterministic and stochastic regimes.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
BEAM: Binary Expert Activation Masking for Dynamic Routing in MoE
arXiv:2605.14438v1 Announce Type: new
Abstract: Mixture-of-Experts (MoE) architectures enhance the efficiency of large language models by activating only a subset of experts per token. However, standard MoE employs a fixed Top-K routing strategy, leading to redundant computation and suboptimal inference latency. Existing acceleration methods either require costly retraining with architectural changes or suffer from severe performance drop at high sparsity due to train-inference mismatch. To address these limitations, we propose BEAM (Binary Expert Activation Masking), a novel method that learns token-adaptive expert selection via trainable binary masks. With a straight-through estimator and an auxiliary regularization loss, BEAM induces dynamic expert sparsity through end-to-end training while maintaining model capability. We further implement an efficient custom CUDA kernel for BEAM, ensuring seamless integration with the vLLM inference framework. Experiments show that BEAM retains over 98\% of the original model's performance while reducing MoE layer FLOPs by up to 85\%, achieving up to 2.5$\times$ faster decoding and 1.4$\times$ higher throughput, demonstrating its effectiveness as a practical, plug-and-play solution for efficient MoE inference.
Fonte: arXiv cs.AI
Vision • Score 85
Vision-Based Runtime Monitoring under Varying Specifications using Semantic Latent Representations
arXiv:2605.13923v1 Announce Type: new
Abstract: We study certified runtime monitoring of past-time signal temporal logic (ptSTL) from visual observations under partial observability. The monitor must infer safety-relevant quantities from images and provide finite-sample guarantees, while being \emph{reusable}: once trained and calibrated, it should certify any formula in a target fragment without per-formula retraining. For fragments induced by a finite dictionary of temporal atoms, we prove that the \emph{semantic basis}, the vector of atom robustness scores, is the minimum prediction target within the class of monotone, 1-Lipschitz reusable interfaces: any formula is evaluated by a deterministic decoder derived from the parse tree, and a single conformal calibration pass certifies the entire fragment with no union bound. We also introduce a \emph{rolling prediction monitor} that predicts only current predicate values and reconstructs temporal history online; this is easier to learn but grows conservative at long horizons. On a pedestrian-crossroad benchmark, rolling achieves tighter certified bounds at short horizons while the semantic-basis monitor is up to 4-times tighter at long horizons. We validate the presented monitors on real-world Waymo driving data, where both monitors satisfy the conformal coverage guarantee empirically.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
GraphBit: A Graph-based Agentic Framework for Non-Linear Agent Orchestration
arXiv:2605.13848v1 Announce Type: new
Abstract: Agentic LLM frameworks that rely on prompted orchestration, where the model itself determines workflow transitions, often suffer from hallucinated routing, infinite loops, and non-reproducible execution. We introduce GraphBit, an engine-orchestrated framework that defines workflows explicitly and deterministically as a directed acyclic graph (DAG). Unlike prompted orchestration, agents in GraphBit operate as typed functions, while a Rust-based engine governs routing, state transitions, and tool invocation, ensuring reproducibility and auditability. The engine supports parallel branch execution, conditional control flow over structured state predicates, and configurable error recovery. A three-tier memory architecture consisting of ephemeral scratch space, structured state, and external connectors isolates context across stages, preventing cascading context bloat that degrades reasoning in long-running pipelines. Across GAIA benchmark tasks spanning zero-tool, document-augmented, and web-enabled workflows, GraphBit outperforms six existing frameworks, achieving the highest accuracy (67.6 percent), zero framework-induced hallucinations, the lowest latency (11.9 ms overhead), and the highest throughput. Ablation studies demonstrate that each memory tier contributes measurably to performance, with deterministic execution providing the greatest gains on tool-intensive tasks representative of real-world deployments.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
EvolveMem:Self-Evolving Memory Architecture via AutoResearch for LLM Agents
arXiv:2605.13941v1 Announce Type: new
Abstract: Long-term memory is essential for LLM agents that operate across multiple sessions, yet existing memory systems treat retrieval infrastructure as fixed: stored content evolves while scoring functions, fusion strategies, and answer-generation policies remain frozen at deployment. We argue that truly adaptive memory requires co-evolution at two levels: the stored knowledge and the retrieval mechanism that queries it. We present EvolveMem, a self-evolving memory architecture that exposes its full retrieval configuration as a structured action space optimized by an LLM-powered diagnosis module. In each evolution round, the module reads per-question failure logs, identifies root causes, and proposes targeted configuration adjustments; a guarded meta-analyzer applies them with automatic revert-on-regression and explore-on-stagnation safeguards. This closed-loop self-evolution realizes an AutoResearch process: the system autonomously conducts iterative research cycles on its own architecture, replacing manual configuration tuning. Starting from a minimal baseline, the process converges autonomously, discovering effective retrieval strategies including entirely new configuration dimensions not present in the original action space. On LoCoMo, EvolveMem outperforms the strongest baseline by 25.7% relative and achieves a 78.0% relative improvement over the minimal baseline. On MemBench, EvolveMem exceeds the strongest baseline by 18.9% relative. Evolved configurations transfer across benchmarks with positive rather than catastrophic transfer, indicating that the self-evolution process captures universal retrieval principles rather than benchmark-specific heuristics. Code is available at https://github.com/aiming-lab/SimpleMem.
Fonte: arXiv cs.LG
RL • Score 85
Learning to Build the Environment: Self-Evolving Reasoning RL via Verifiable Environment Synthesis
arXiv:2605.14392v1 Announce Type: new
Abstract: We pursue a vision for self-improving language models in which the model does not merely generate problems or traces to imitate, but constructs the environments that train it. In zero-data reasoning RL, this reframes self-improvement from a data-generation loop into an environment-construction loop, where each artifact is a reusable executable object that samples instances, computes references, and scores responses. Whether this vision sustains improvement hinges on a single property: the environments must exhibit stable solve--verify asymmetry, the model must be able to write an oracle once that it cannot reliably execute in natural language on fresh instances. This asymmetry takes two complementary forms. Some tasks are algorithmically hard to reason through but trivial as code: a dynamic program or graph traversal, compiled once, yields unboundedly many calibrated instances. Others are intrinsically hard to solve but easy to verify, like planted subset-sum or constraint satisfaction. Both create a durable gap between proposing and solving that the policy cannot close by gaming the verifier, and it is this gap that keeps reward informative as the learner improves. We instantiate this view in EvoEnv, a single-policy generator, solver method that synthesizes Python environments from ten seeds and admits them only after staged validation, semantic self-review, solver-relative difficulty calibration, and novelty checks. The strongest evidence comes from the already-strong regime: on Qwen3-4B-Thinking, fixed public-data RLVR and fixed hand-crafted environment RLVR reduce the average, while EvoEnv improves it from 72.4 to 74.8, a relative gain of 3.3%. Stable self-improvement, we suggest, depends not on producing more synthetic data, but on models learning to construct worlds whose difficulty stays structurally beyond their own reach.
Fonte: arXiv cs.AI
Applications • Score 85
Agentic AI Ecosystems in Higher Education: A Perspective on AI Agents to Emerging Inclusive, Agentic Multi-Agent AI Framework for Learning, Teaching and Institutional Intelligence
arXiv:2605.14266v1 Announce Type: new
Abstract: Integration of artificial intelligent (AI) agents in higher education is transforming teaching, learning and administrative processes. Although existing AI agents effectively support individual tasks, their implementation remains fragmented and inefficient for handling the complexity of educational institutions. This highlights a significant research gap: the lack of integrated eco-system-level agentic multi-agent AI platform capable of coordinated planning, reasoning, and adaptive decision-making across multiple educational functions. This paper presents a forward-looking perspective on agentic multi-agent AI platform in higher education, consisting interconnected autonomous, goal driven agents that support learning, teaching, and institutional operations. It addresses timely and critical questions: Can agentic AI represent the next generation of intelligent systems in tertiary education? Can they collectively support seamless coordinated operations across teaching, learning and administrative support? To what extent can such systems foster inclusive and equitable learning for diverse learners with special educational needs? To ground this perspective, a thematic analysis of existing literature identifies four dominant themes: task-specific fragmented AI tools, the transition from single-agent to multi-agent systems, limited cross-functional integration, and insufficient focus on inclusivity and accessibility. Findings reveal a clear gap between current AI implementations and the needs of holistic, learner-centered educational ecosystem. The paper synthesizes challenges and outlines future research directions for scalable human-aligned, and inclusive agentic AI platform. The significant contribution is the incorporation of inclusive learning perspectives, highlighting how coordinated agentic multi-agent platform can support diverse learners through adaptive, multimodal interventions.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Know When To Fold 'Em: Token-Efficient LLM Synthetic Data Generation via Multi-Stage In-Flight Rejection
arXiv:2605.14062v1 Announce Type: new
Abstract: While synthetic data generation with large language models (LLMs) is widely used in post-training pipelines, existing approaches typically generate full outputs before applying quality filters, leading to substantial token waste on samples that are ultimately discarded. To address this, we propose Multi-Stage In-Flight Rejection (MSIFR), a lightweight, training-free framework that detects and terminates low-quality generation trajectories at intermediate checkpoints before they reach full completion. MSIFR decomposes the generation process into sequential stages and applies fast rule-based validators to identify arithmetic inconsistencies, hallucination patterns, and formatting violations, enabling early rejection of faulty samples. We formalize in-flight rejection as a sequential decision process and show that any non-trivial discard policy reduces expected token consumption, with stage-wise savings increasing when rejection occurs earlier in the generation pipeline. We further demonstrate that conditional utility estimates form a martingale, ensuring that early, in-flight rejection does not bias the expected utility of retained samples. Across five instruction-tuned models and seven reasoning benchmarks, MSIFR reduces token consumption by 11%-77% as a standalone method, and up to 78.2% when combined with early-exit methods, while preserving or improving evaluation accuracy. These results confirm that MSIFR provides a practical mechanism for improving the efficiency of LLM-based synthetic data generation without additional training or architectural changes.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
CSI-JEPA: Towards Foundation Representations for Ubiquitous Sensing with Minimal Supervision
arXiv:2605.14171v1 Announce Type: new
Abstract: Channel state information (CSI) provides a widely available sensing modality for human and environment perception, but existing CSI sensing models usually rely on task-specific supervised training and require substantial labeled data for each task, device, user, or environment. This limits their scalability in practical deployments where unlabeled CSI is abundant but labeled data is costly to collect. In this paper, we present CSI-JEPA, a self-supervised predictive representation learning framework for label-efficient, multi-task Wi-Fi sensing. CSI-JEPA learns reusable temporal-spectral representations from unlabeled CSI samples by predicting latent features of masked channel regions from visible context. To better match the physical structure of CSI, CSI-JEPA tokenizes channel-response amplitude windows along the time and subcarrier dimensions. It then introduces a channel variation-aware masking strategy that samples predictive targets from regions with stronger local temporal and subcarrier-domain variations. After pretraining, the encoder is frozen and used as a backbone, with lightweight task-specific adapters added for downstream sensing tasks. We evaluate CSI-JEPA on seven real-world Wi-Fi sensing tasks spanning diverse objectives and deployment settings. The results show that CSI-JEPA improves downstream sensing performance over competitive baselines, achieving up to 10.64 percentage points mean accuracy gain over state-of-the-art supervised Transformer and matched-budget label savings of up to 98.0%.
Fonte: arXiv cs.LG
NLP/LLMs • Score 75
Pitfalls of Unlabeled Disagreement-Based Drift Detection in Streaming Tree Ensembles
arXiv:2605.12803v1 Announce Type: new
Abstract: Detecting concept drift in high-speed data streams remains challenging, particularly when models must operate on unlabeled data and avoid false alarms caused by benign shifts. While disagreement-based uncertainty has shown promise in neural networks, its adaptation to ensembles of incremental decision trees (IDTs) remains largely unexplored. We investigate this approach by constructing batch-specific disagreement measures via label flipping in ensemble members and evaluating their effectiveness for drift detection in tabular data streams. Our experiments show that, although this method performs well in ensembles of multi-layer perceptrons (MLPs), it consistently underperforms loss-based detectors when applied to IDTs. We attribute this behavior to the intrinsic rigidity of IDTs: learning primarily through structural expansion, with limited parameter adaptation, restricts model plasticity and prevents disagreement from reliably reflecting learning potential. Recent work on restructuring IDTs using their intrinsic decomposition into non-overlapping rules offers a promising direction for improving adaptability.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
CAWI: Copula-Aligned Weight Initialization for Randomized Neural Networks
arXiv:2605.12580v1 Announce Type: new
Abstract: Randomized neural networks (RdNNs) enable efficient, backpropagation-free training by freezing randomly initialized input-to-hidden weights, which permits a closed-form solution for the output layer. However, conventional random initialization is blind to inter-feature dependence, ignoring correlations, asymmetries, and tail dependence in the data, which degrades conditioning and predictive performance. To the best of our knowledge, this limitation remains unaddressed in the RdNN literature. To close this gap, we propose CAWI (Copula-Aligned Weight Initialization), a framework that draws input-to-hidden weights from a data-fitted copula that matches empirical dependence, ensuring the frozen projections respect inter-feature dependence without sacrificing the closed-form solution. CAWI (i) maps each feature to the unit interval using empirical CDFs, (ii) fits a multivariate copula that captures rank-based dependence among features, and (iii) samples each weight column w_j from the fitted copula and applies a fixed inverse marginal transform to set scale. The objective, solver, and "freeze-once" paradigm remain unchanged; only the sampling law for W becomes dependence-aware. For dependence modeling, we consider two copula families: elliptical (Gaussian, t) and Archimedean (Clayton, Frank, Gumbel). This enables CAWI to handle diverse dependence, including tail dependence. We evaluate CAWI across 83 diverse classification benchmarks (binary and multiclass) and two biomedical datasets, BreaKHis and the Schizophrenia dataset, using standard shallow and deep RdNN architectures. CAWI consistently delivers significant improvements in predictive performance over conventional random initialization. Code is available at: https://github.com/mtanveer1/CAWI
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Rescaled Asynchronous SGD: Optimal Distributed Optimization under Data and System Heterogeneity
arXiv:2605.13434v1 Announce Type: cross
Abstract: Asynchronous stochastic gradient descent (ASGD) is a standard way to exploit heterogeneous compute resources in distributed learning: instead of forcing fast workers to wait for slow ones, the server updates the model whenever a gradient arrives. Vanilla ASGD applies each arriving gradient with the same weight. When local data distributions are heterogeneous, this becomes problematic: faster workers contribute more updates, and we show theoretically that the method is biased toward a frequency-weighted average of the local objectives rather than the desired global objective. Existing remedies typically move away from the simple ASGD template by introducing gathering phases, buffering, or extra memory. We show that this is unnecessary. Keeping the standard ASGD mechanism, we recover the correct objective by rescaling worker-specific stepsizes in proportion to their computation times, so that each worker contributes the same aggregate learning rate over a cycle. In the non-convex setting, under smoothness and bounded heterogeneity assumptions, we prove that the resulting method, Rescaled ASGD, converges to stationary points of the correct global objective in the fixed-computation model. Its time complexity matches the known lower bound in the leading term, while the effects of staleness and data heterogeneity appear only in lower-order terms. Experiments confirm that the method converges to the correct objective and is competitive with state-of-the-art baselines.
Fonte: arXiv stat.ML
RL • Score 92
D-VLA: A High-Concurrency Distributed Asynchronous Reinforcement Learning Framework for Vision-Language-Action Models
arXiv:2605.13276v1 Announce Type: new
Abstract: The rapid evolution of Embodied AI has enabled Vision-Language-Action (VLA) models to excel in multimodal perception and task execution. However, applying Reinforcement Learning (RL) to these massive models in large-scale distributed environments faces severe systemic bottlenecks, primarily due to the resource conflict between high-fidelity physical simulation and the intensive VRAM/bandwidth demands of deep learning. This conflict often leaves overall throughput constrained by execution-phase inefficiencies. To address these challenges, we propose D-VLA, a high-concurrency, low-latency distributed RL framework for large-scale embodied foundation models. D-VLA introduces "Plane Decoupling," physically isolating high-frequency training data from low-frequency weight control to eliminate interference between simulation and optimization. We further design a four-thread asynchronous "Swimlane" pipeline, enabling full parallel overlap of sampling, inference, gradient computation, and parameter distribution. Additionally, a dual-pool VRAM management model and topology-aware replication resolve memory fragmentation and optimize communication efficiency. Experiments on benchmarks like LIBERO show that D-VLA significantly outperforms mainstream RL frameworks in throughput and sampling efficiency for billion-parameter VLA models. In trillion-parameter scalability tests, our framework maintains exceptional stability and linear speedup, providing a robust system for high-performance general-purpose embodied agents.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
Yield Curves Dynamics Using Variational Autoencoders Under No-arbitrage
arXiv:2605.12764v1 Announce Type: cross
Abstract: This paper introduces a physics-informed generative framework that resolves the fundamental conflict between the statistical flexibility of deep learning and the rigorous theoretical constraints of fixed-income modeling. We demonstrate that standard generative models and unconstrained statistical extrapolations suffer from "manifold collapse" and severe arbitrage violations when forecasting term structures across diverse macroeconomic regimes. To overcome this, we propose a two-stage architecture. First, a Student-t Conditional Variational Autoencoder with Dynamic Level Injection (CVAEsT+LS) extracts a robust, heavy-tailed term structure manifold, effectively decoupling macroeconomic shape dynamics from absolute base rates. Second, the latent dynamic evolution is governed by a continuous-time Neural Stochastic Differential Equation (SDE) strictly penalized by a No-Arbitrage Partial Differential Equation (PDE). Empirical results across multiple sovereign currencies (USD, GBP, JPY) confirm that our synergistic approach drastically reduces out-of-sample forecasting errors -- achieving an exceptional 6.58 bps Mean Tenor RMSE -- and successfully overcomes the massive parallel drift and zero-lower-bound violations exhibited by the classical HJM model in extreme environments. Furthermore, through phase space vector field analysis, we demonstrate the model's superior capability in unsupervised macroeconomic regime detection and high-quality continuous-time scenario generation. Ultimately, this research provides a highly scalable, mathematically sound evolutionary engine for term structure modeling.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Runtime Monitoring of Perception-Based Autonomous Systems via Embedding Temporal Logic
arXiv:2605.12651v1 Announce Type: new
Abstract: Runtime monitoring of autonomous systems traditionally relies on mapping continuous sensor observations to discrete logical propositions defined over low-dimensional state variables. This abstraction breaks down in perception-driven settings, where such mappings require additional learned modules that are often computationally expensive, brittle, and semantically misaligned. In this work, we propose Embedding Temporal Logic (ETL), a temporal logic that performs monitoring directly in learned embedding spaces. ETL defines predicates through distances between observed embeddings and target embeddings derived from reference observations. This formulation allows specifications to capture high-level perceptual concepts, such as similarity to visual goals or avoidance of semantic regions, that are difficult or impossible to express using traditional predicates. By composing these predicates with temporal operators, ETL naturally expresses temporally extended and sequential perceptual behaviors. We introduce ETL monitors for evaluating specifications over bounded embedding traces, along with a conformal calibration procedure that provides reliable and safety-oriented predicate evaluation. We evaluate our approach across multiple manipulation environments to show that ETL achieves strong empirical agreement with ground-truth semantics, including accurate monitoring of temporally composed behaviors.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
IdeaForge: A Knowledge Graph-Grounded Multi-Agent Framework for Cross-Methodology Innovation Analysis and Patent Claim Generation
arXiv:2605.13311v1 Announce Type: new
Abstract: Current AI-assisted innovation systems typically apply a single ideation methodology (such as TRIZ or Design Thinking) using sequential prompt-based workflows that do not preserve intermediate reasoning structure. As a result, insights generated across methodologies remain fragmented, limiting traceability, synthesis, and systematic evaluation of novelty. We present IdeaForge, a knowledge graph-grounded multi-agent framework for innovation analysis and patent claim generation. IdeaForge integrates multiple innovation methodologies (TRIZ, Design Thinking, and SCAMPER) through specialist agents operating over a persistent FalkorDB knowledge graph. Each agent contributes structured entities and relationships representing contradictions, inventive principles, user needs, transformations, analogies, and candidate claims. The central contribution of IdeaForge is a cross-methodology convergence mechanism implemented through graph-based claim linkage. Claims independently supported by multiple methodologies are connected using CONVERGENT relationships, enabling identification of high-confidence innovation candidates through graph traversal. A downstream patent drafting agent generates structured patent drafts grounded in convergent claim subgraphs, reducing reliance on unconstrained language model generation. An InnovationScore formula ranks claims by convergent support, methodology diversity, claim strength, and prior art challenge count. We describe the graph schema, agent architecture, convergence detection pipeline, and patent synthesis workflow. Experiments on a legal technology use case demonstrate that graph-grounded multi-methodology synthesis produces more diverse and traceable innovation candidates compared to single-methodology baselines. We discuss implications for computational creativity, explainable AI-assisted invention, and graph-native innovation systems.
Fonte: arXiv cs.AI
Multimodal • Score 85
Towards Robust Federated Multimodal Graph Learning under Modality Heterogeneity
arXiv:2605.12584v1 Announce Type: new
Abstract: Recently, multimodal graph learning (MGL) has garnered significant attention for integrating diverse modality information and structured context to support various network applications. However, real-world graphs are often isolated due to data-sharing limitations across multiple parties, and their modalities are frequently incomplete. This highlights an urgent need to develop a robust federated approach. However, we find that existing methods remain insufficient. On the one hand, centralized MGL methods that handle missing modalities overlook the knowledge sharing and generalization in federated scenarios. On the other hand, while federated MGL methods have become increasingly mature, they primarily target non-graph data. Based on these technologies, we identify a two-stage pipeline wherein client-side completion reconstructs missing modalities, and server-side aggregation integrates the client-updated parameters of both the modality generator and the backbone models. Although this serves as a general solution, we identify two primary challenges in achieving greater robustness: (1) Topology-Isolated Local Completion: Client-side modality generation struggles to effectively leverage global semantics. (2) Reliability-Imbalanced Global Aggregation: Server-side multi-party collaboration is hindered by client updates with varying modality availability and recovery reliability. To address these challenges, we propose \textsc{FedMPO}, which utilizes topology-aware cross-modal generation to recover missing features using comprehensive graph context, missing-aware expert routing to locally filter out noisy recovered signals, and reliability-aware aggregation to appropriately down-weight unreliable updates. Extensive experiments on 3 tasks across 6 datasets demonstrate that FedMPO outperforms baselines, achieving performance gains of up to 4.10% and 5.65% in high-missing and non-IID settings.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Bayesian Model Merging
arXiv:2605.12843v1 Announce Type: new
Abstract: Model merging aims to combine multiple task-specific expert models into a single model without joint retraining, offering a practical alternative to multi-task learning when data access or computational budget is limited. Existing methods, however, face two key limitations: (1) they overlook the valuable inductive bias of strong anchor models and estimate the merged weights from scratch, and (2) they rely on a shared hyperparameter setting across different modules of the network, lacking a global optimization strategy. This paper introduces Bayesian Model Merging (BMM), a plug-and-play bi-level optimization framework, where the inner level formulates the model merging as an activation-based Bayesian regression under a strong prior induced by an anchor model, yielding an efficient closed-form solution; and the outer level leverages a Bayesian optimization procedure to search module-specific hyperparameters globally based on a small validation set. Furthermore, we reveal a key alignment between activation statistics and task vectors, enabling us to derive a data-free variant of BMM that estimates the Gram matrix for regression without any auxiliary data. Across extensive benchmarks, including up to 20-task merging in vision and 5-task merging in language, BMM consistently outperforms all plug-and-play anchor baselines (e.g., TA, WUDI-Merging, and TSV). In particular, on the ViT-L/14 benchmark for 8-task merging, a single merged model reaches 95.1, closely matching the average performance of eight task-specific experts (95.8).
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
UFO: A Domain-Unification-Free Operator Framework for Generalized Operator Learning
arXiv:2605.12700v1 Announce Type: new
Abstract: Neural operators have become an effective framework for learning mappings between function spaces, yet most existing architectures realize operators within a single representational domain, such as physical, spectral, or latent space. In this work, we introduce UFO (Domain-Unification-Free Operator), a cross-domain neural operator framework that realizes operators through adaptive, jointly conditioned interactions among representations defined on distinct domains. UFO enables discretization decoupling: the input function can be observed at resolutions or locations different from those used during training, while the solution can be queried at arbitrary output resolutions. Across four complementary benchmarks covering discontinuous inputs, irregular sampling with spectral mismatch, nonlinear dynamics, and stochastic high-frequency fields, UFO delivers accurate, robust, and physically coherent predictions under distribution shifts. These results establish cross-domain, phase-modulated realization as a powerful framework for discretization-decoupled neural operator learning.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Multi-Agent Systems in Emergency Departments: Validation Study on a ED Digital Twin
arXiv:2605.13345v1 Announce Type: new
Abstract: Emergency departments (ED) face challenges in patient care and resource management. We propose to explore optimization strategies in a realistic and flexible model and develop a hybrid Discrete Event Simulation (DES) and Agent-Based Model (ABM) simulating highly configurable ED environments. We specifically focus on the validation of the modeling approach. We derive configurations for ED sizes, patient load, and staffing from real-world studies. We then validate the model expressivity by matching its key performance indicators and metrics with their values known from literature. We proceed by implementing scientifically established and practice-proven resource optimization strategies. Comparing the documented real-world outcomes with our model's results demonstrates that the DES-ABM based simulation can effectively replicate real-world ER dynamics under interventions. We lastly integrate a Proof-of-Concept multi-agent system (MAS) that can autonomously explore resource allocation strategies within the simulated ER environment based on a temporal ledger of ED event records. This modular DES-ABM-MAS framework offers a powerful tool to explore resource optimization strategies in emergency departments.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
It's not the Language Model, it's the Tool: Deterministic Mediation for Scientific Workflows
arXiv:2605.13245v1 Announce Type: new
Abstract: Language models can produce convincing scientific analyses, but repeated generations on the same data do not guarantee the same result. A researcher may regenerate an identical query and receive a different fit, a different peak position or a different analysis procedure, without an obvious way to decide which output to trust. We propose typed mediation, a pattern in which the model orchestrates deterministic tools rather than generating analytical code. Each tool encodes one researcher's exact procedure for one instrument, ported through structured interviews. The model selects which tool to call and with what parameters. The tool produces the result. Regeneration does not change it. We evaluate this claim by running the same photoluminescence analysis on four platforms, including three commercial foundation models, four times each with the same prompt. The typed tool produces identical results across all runs. The commercial platforms either vary in numerical output and analytical methodology across runs, or fail to produce valid results on the task. We deploy this pattern on two instruments serving users over approximately six months, with very positive user feedback. Both cases are very challenging: they involve proprietary binary formats and per-seat licensed software, which force the tool to remain on local infrastructure alongside the data and the instrument it operates. We argue that deployment topology is not just a preference, but a structural requirement of scientific tool mediation. The result is a practical pattern for deploying language models in scientific workflows where reproducibility is mandatory, reducing analysis time from weeks to minutes while guaranteeing identical outputs across runs.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Scale-Gest: Scalable Model-Space Synthesis and Runtime Selection for On-Device Gesture Detection
arXiv:2605.12506v1 Announce Type: new
Abstract: Realizing on-device ML-based gesture detection under tight real-time performance, energy and memory constraints is challenging, especially when considering mobile devices with varying battery-power levels. Existing EdgeAI deployments typically rely on a single fixed detector, limiting optimization opportunities. We present Scale-Gest, a novel run-time adaptive gesture detection framework that expands the detector space into a dense family of tiny-YOLO architectures. We introduce multiple novel device-calibrated ACE (Accuracy-Complexity-Energy) profiles by analyzing different model-resolution-stride operating points. A lightweight run-time controller selects an appropriate ACE mode under user-defined and battery constraints, while a motion-aware hand-gesture-tracking ROI gate crops the input for reduced complexity detection. To evaluate performance of our system in real-world car driving scenarios, we introduce a temporally-annotated Driver Simulated Gesture (DSG-18) dataset. Scale-Gest maintains event-level F1 while significantly reducing energy and latency compared to single-detector approaches. On a battery-powered laptop running gesture streams, our ACE controller reduces per-frame energy by 4x (from 6.9 mJ to 1.6 mJ) while maintaining high gesture-detection performance (event-level F1 = 0.8-0.9) and low mean latency (6 ms).
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
An Agentic LLM-Based Framework for Population-Scale Mental Health Screening
arXiv:2605.13046v1 Announce Type: new
Abstract: Mental health disorders affect millions worldwide, and healthcare systems are increasingly overwhelmed by the volume of clinical data generated from electronic records, telemedicine platforms, and population-level screening programs. At the same time, the emergence of novel AI-based approaches in healthcare calls for intelligent frameworks capable of processing domain-specific unstructured clinical information while adapting to patient-specific needs. This paper proposes an agentic framework for building robust LLM-based pipelines, where each stage is encapsulated as a LangChain agent governed by explicit policies and proxy-guided evaluation. Stages are incrementally locked once validated, ensuring that later adaptations cannot overwrite configurations without demonstrated improvement. The proposed framework evolves from feature-level exploration, through proxy-based tuning and freeze/rollback mechanisms, to full orchestration by an Orchestrator Agent that coordinates preprocessing, retrieval, selection, diversity, threshold optimization, and decoding. A proof-of-concept in transcript-based depression detection demonstrates that the framework converges to stable configurations, such as cosine similarity, dynamic Top-k, and threshold 0.75, while controlling evaluation costs and avoiding regressions. These results highlight the potential of agentic AI to enable population-level mental health screening over large clinical datasets, addressing critical challenges in trustworthiness, reproducibility, and adaptability required in healthcare environments.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
When Should an AI Workflow Release? Always-Valid Inference for Black-Box Generate-Verify Systems
arXiv:2605.12947v1 Announce Type: new
Abstract: LLM-enabled AI workflows increasingly produce outputs through iterative generate-evaluate-revise loops. Each iteration can improve the candidate, but it also creates a release decision: when to stop and output the current result? This raises a statistical challenge because deployment-time evaluator scores are adaptively generated and repeatedly monitored, yet the likelihood models or exchangeability assumptions typically used for calibration are unavailable. We propose an always-valid release wrapper for existing generator-evaluator pipelines. The wrapper builds a hard-negative reference pool of high-scoring failures, calibrates deployment-time evaluator scores against this pool, and accumulates the resulting evidence with an e-process. This separates two roles: the reference pool turns black-box scores into conservative evidence, while the e-process provides validity under optional stopping. In theory, we show that a conservative reference pool yields finite-sample control of the probability of releasing on infeasible tasks, that is, tasks for which the given workflow is not capable of producing a reliable solution. We also characterize conditions under which the same conservative rule still achieves nontrivial release on feasible tasks. In an MBPP+ coding-agent case study, the wrapper reduces premature incorrect release relative to baseline stopping rules while still releasing on tasks for which the workflow repeatedly accumulates moderate supporting evidence.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
An Agentic AI Framework with Large Language Models and Chain-of-Thought for UAV-Assisted Logistics Scheduling with Mobile Edge Computing
arXiv:2605.13221v1 Announce Type: new
Abstract: In cloud manufacturing, unmanned aerial vehicles (UAVs) can support both product collection and mobile edge computing (MEC). This joint operation forms a hybrid scheduling problem, where physical logistics decisions are coupled with computational task scheduling. In this paper, UAVs collect finished products from manufacturing stations and transport them back to a central depot. Meanwhile, computational tasks generated by industrial sensor devices at these stations are processed locally, at UAVs, or offloaded via UAVs to the cloud. This coupling makes the problem challenging. A UAV can provide MEC services only during its service window at a station, so routing decisions directly determine when UAV-assisted offloading is available. Routing decisions also affect the UAV energy budget and the availability of onboard computing and communication resources for computational task execution under task deadline constraints. To address this, we propose an agentic-AI-assisted optimization framework with two components. First, we develop an agentic AI that combines large language models, retrieval-augmented generation, and chain-of-thought reasoning to translate user input into an interpretable mathematical formulation for the hybrid scheduling problem. Second, we design a hierarchical deep reinforcement learning approach based on proximal policy optimization (PPO), where the upper layer learns UAV routing and the lower layer optimizes per-slot task execution and resource allocation. Simulation results show that the proposed framework yields more consistent formulations, while the hierarchical PPO achieves full product collection in 99.6% of the last 500 episodes and maintains a 100% deadline satisfaction rate, with more stable performance than the advantage actor-critic approach.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Digital Twins as Synthetic Controls in Single-Arm Trials
arXiv:2605.12832v1 Announce Type: cross
Abstract: Single-arm trials are an important study design for evaluating drug efficacy and safety without enrolling patients into a control arm. Although they do not provide the gold-standard evidence of randomized controlled trials, they are increasingly used in clinical development as they offer an efficient, ethical, and practical alternative. A wide variety of approaches can be used to construct control comparators and estimate treatment effects, from fixed comparators informed by clinical knowledge to data-based and model-based patient-level comparators, also known as synthetic controls. Powerful and flexible machine learning models can allow outcome-model-based synthetic controls to overcome key limitations of direct data-based approaches, yield more robust estimates of treatment effects, and provide a principled way to incorporate corrections or encode additional assumptions when external data are not directly comparable. In this work, we argue that outcome-model-based synthetic control arms are an important tool for single-arm trials. We focus on digital twins, personalized predictions of disease progression generated from machine learning models trained on historical datasets, which naturally leverage these flexible approaches. We review doubly robust estimators, present power and sample size formulas, and discuss trade-offs in selecting historical data for training and analysis. We also outline practical considerations for deploying digital twins within the framework of recent FDA draft guidance on the use of artificial intelligence in drug development. Finally, we reanalyze data from trials in amyotrophic lateral sclerosis and Huntington's disease to demonstrate the proposed methods.
Fonte: arXiv stat.ML
Vision • Score 85
EgoForce: Robust Online Egocentric Motion Reconstruction via Diffusion Forcing
arXiv:2605.13041v1 Announce Type: new
Abstract: With recent advances in embodied agents and AR devices, egocentric observations are readily available as input for real-world interactive online applications. However, egocentric viewpoints can only sporadically observe hands, in addition to the estimated head trajectory. We propose EgoForce, an online framework for reconstructing long-term full-body motion from noisy egocentric input. While existing generative frameworks can robustly handle noisy and sparse measurements, they assume a fixed-length observation window is available and are thus not suitable for real-time applications. Faster inference often relies on autoregressive prediction, sacrificing robustness. In contrast, we adopt a diffusion-based method with a temporally asymmetric noise schedule inspired by Diffusion Forcing. Specifically, our approach models temporally evolving uncertainty and incrementally denoises states as new streaming observations arrive. Combined with a noise-robust imputation strategy, EgoForce progressively generates stable and coherent full-body motion under strict causal constraints. Experiments demonstrate that our online framework outperforms existing online and offline methods, enabling long-horizon, full-body motion reconstruction in challenging egocentric scenarios.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
DIVER:Diving Deeper into Distilled Data via Expressive Semantic Recovery
arXiv:2605.12649v1 Announce Type: new
Abstract: Dataset distillation aims to synthesize a compact proxy dataset that is unreadable or non-raw from the original dataset for privacy protection and highly efficient learning. However, previous approaches typically adopt a single-stage distillation paradigm, which suffers from learning specific patterns that overfit on a prior architecture, consequently suppressing the expression of semantics and leading to performance degradation across heterogeneous architectures. To address this issue, we propose a novel dual-stage distillation framework called ${\textbf{DIVER}}$, which leverages the pre-trained diffusion model to dive deeper into $\textbf{DI}$stilled data $\textbf{V}$ia $\textbf{E}$xpressive semantic $\textbf{R}$ecovery, an entire process of semantic inheritance, guidance, and fusion. Semantic inheritance distills high-level semantics of abstract distilled images into the latent space to filter out architecture-specific ``noise" and retain the intrinsic semantics. Furthermore, semantic guidance improves the preservation of the original semantics by directing the reverse procedure. Finally, semantic fusion is designed to provide semantic guidance only during the concrete phase of the reverse process, preventing semantic ambiguity and artifacts while maintaining the guidance information. Extensive experiments validate the effectiveness and efficiency of DIVER in improving classical distillation techniques and significantly improving cross-architecture generalization, requiring processing time comparable to raw DiT on ImageNet (256$\times$256) with only 4 GB of GPU memory usage. Code is available: https://github.com/einsteinxia/DIVER.
Fonte: arXiv cs.CV
Theory/Optimization • Score 85
Discrete MeanFlow: One-Step Generation via Conditional Transition Kernels
arXiv:2605.12805v1 Announce Type: new
Abstract: MeanFlow enables one-step generation in continuous spaces by learning an average velocity over a time interval rather than the instantaneous velocity field of flow matching. However, discrete state spaces do not have smooth trajectories or spatial derivatives, so the continuous formulation does not directly apply. We introduce Discrete MeanFlow, which replaces the motion of a point with the transport of probability mass over finite states. Our key object is the conditional transition kernel of a continuous-time Markov chain (CTMC), from which we define a mean discrete rate that measures the average change in transition probability over a time interval. We prove a Discrete MeanFlow identity that relates this finite-interval rate to the instantaneous CTMC generator at the endpoint, with the Kolmogorov forward equation replacing the spatial chain rule of continuous MeanFlow. Based on this identity, we parameterize the transition kernel directly using a boundary-by-construction design that guarantees valid probability outputs and exact boundary conditions without auxiliary losses. Since the learned kernel is itself a probability distribution, generation reduces to a single forward pass followed by one categorical draw meaning no iterative denoising, ODE integration, or multi-step refinement is required. We validate the framework on exact finite-state Markov chains, where the learned kernel recovers the analytical ground truth to high precision, and on factorized synthetic sequence generation tasks with varying alphabet sizes and sequence lengths.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Predicting Channel Closures in the Lightning Network with Machine Learning
arXiv:2605.12759v1 Announce Type: new
Abstract: The Lightning Network (LN) is a second-layer protocol for Bitcoin designed to enable fast and cost-efficient off-chain transactions. Channels in the LN can be closed either by mutual agreement or unilaterally through a forced closure, which locks the involved capital for an extended period and degrades network reliability. In this paper, we study the problem of predicting channel closure types from publicly available gossip data, framing it as a temporal link classification task over the evolving channel graph. We construct a dataset spanning over two years of LN activity and benchmark a range of machine learning approaches, from MLPs to temporal graph neural networks and spectral encodings. Our experiments reveal that the dominant predictive signals are temporal and behavioural, namely how recently each endpoint was active and the per-node history of past closures, while the surrounding network topology provides no additional benefit. We find that a simple MLP operating on edge-level features, node-level event counts, and temporal patterns outperforms all graph-based approaches, and discuss how the inherent privacy of the LN, where critical information such as channel balances and payment flows remains hidden, fundamentally limits the predictability of closures from gossip data alone. We publicly release the dataset and code at https://github.com/AmbossTech/ln-channel-closure-prediction to encourage further research on this practically relevant task.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
MorphOPC: Advancing Mask Optimization with Multi-scale Hierarchical Morphological Learning
arXiv:2605.12528v1 Announce Type: new
Abstract: As feature sizes shrink to the nanometer scale, accurately transferring circuit patterns from photomasks to silicon wafers becomes increasingly challenging. Optical proximity correction (OPC) is widely used to ensure pattern fidelity and manufacturability. Recent generative mask optimization models based on encoder-decoder architecture can synthesize near-optimal masks, serving as fast machine learning (ML) surrogates for traditional OPC. However, these models often fail to capture the geometric transformations from target layouts to mask patterns, leading to suboptimal quality. In this work, we formulate mask generation as a sequence of morphological operations on local layout features and propose \textit{MorphOPC}, a multi-scale hierarchical model with neural morphological modules to learn these transformations. Experiments on edge-based OPC and ILT benchmarks across metal and via layers show that \textit{MorphOPC} consistently outperforms state-of-the-art methods, achieving higher printing fidelity and lower manufacturing cost, demonstrating strong potential for scalable mask optimization.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Orthrus: Memory-Efficient Parallel Token Generation via Dual-View Diffusion
arXiv:2605.12825v1 Announce Type: new
Abstract: We introduce Orthrus, a simple and efficient dual-architecture framework that unifies the exact generation fidelity of autoregressive Large Language Models (LLMs) with the high-speed parallel token generation of diffusion models. The sequential nature of standard autoregressive decoding represents a fundamental bottleneck for high-throughput inference. While diffusion language models attempt to break this barrier via parallel generation, they suffer from significant performance degradation, high training costs, and a lack of rigorous convergence guarantees. Orthrus resolves this dichotomy natively. Designed to seamlessly integrate into existing Transformers, the framework augments a frozen LLM with a lightweight, trainable module to create a parallel diffusion view alongside the standard autoregressive view. In this unified system, both views attend to the exact same high-fidelity Key-Value (KV) cache; the autoregressive head executes context pre-filling to construct accurate KV representations, while the diffusion head executes parallel generation. By employing an exact consensus mechanism between the two views, Orthrus guarantees lossless inference, delivering up to a 7.8x speedup with only an O(1) memory cache overhead and minimal parameter additions.
Fonte: arXiv cs.LG
Evaluation/Benchmarks • Score 85
Do Androids Dream of Breaking the Game? Systematically Auditing AI Agent Benchmarks with BenchJack
arXiv:2605.12673v1 Announce Type: new
Abstract: Agent benchmarks have become the de facto measure of frontier AI competence, guiding model selection, investment, and deployment. However, reward hacking, where agents maximize a score without performing the intended task, emerges spontaneously in frontier models without overfitting. We argue that benchmarks must be secure by design. From past incidents of reward hacks, we derive a taxonomy of eight recurring flaw patterns and compile them into the Agent-Eval Checklist for benchmark designers. We condense the insights into BenchJack, an automated red-teaming system that drives coding agents to audit benchmarks and identify possible reward-hacking exploits in a clairvoyant manner. Moreover, we extend BenchJack to an iterative generative-adversarial pipeline that discovers new flaws and patches them iteratively to improve benchmark robustness. We apply BenchJack to 10 popular agent benchmarks spanning software engineering, web navigation, desktop computing, and terminal operations. BenchJack synthesizes reward-hacking exploits that achieve near-perfect scores on most of the benchmarks without solving a single task, surfacing 219 distinct flaws across the eight classes. Moreover, BenchJack's extended pipeline reduces the hackable-task ratio from near 100% to under 10% on four benchmarks without fatal design flaws, fully patching WebArena and OSWorld within three iterations. Our results show that evaluation pipelines have not internalized an adversarial mindset, and that proactive auditing could help close the security gap for the fast-paced benchmarking space.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Neurodata Without Boredom: Benchmarking Agentic AI for Data Reuse
arXiv:2605.12808v1 Announce Type: new
Abstract: Neuroscience data are highly fragmented across labs, formats, and experimental paradigms, and reuse often requires substantial manual effort. A persistent roadblock to data reuse and integration is the need to decipher bespoke and diverse data formatting choices. Common data formats have been proposed in response, but the field continues to struggle with a fundamental tension: formats flexible enough to accommodate diverse experiments are rarely descriptive enough to be self-explanatory, and sufficiently descriptive formats demand detailed documentation and curation effort that few labs can sustain. Agentic AI is a natural candidate to solve this problem: LLMs read code and text faster and with sustained attention to the low-level details humans tend to skim over. To measure how well agentic AI performs on this task, we selected eight recent papers studying large-scale mouse neural population recordings that shared both data and code, spanning diverse recording modalities, behavioral paradigms, and dataset formats (e.g., NWB, specialized APIs, and general-purpose Python or MATLAB files). We provided agents with the data, code, and paper, and prompted them to load, understand, and reformat the data for a common downstream task: training a decoder from neural activity to task or behavioral variables. General-purpose coding agents commonly used by scientists performed well on each sub-task, but rarely strung together a fully error-free end-to-end solution. We characterize the types of mistakes agents made and the dataset properties that elicited them, and propose data-sharing best practices for the agentic-AI era. We further find that agents-as-judges are unreliable at catching errors, especially without ground-truth references, so interactive, human-in-the-loop coding remains necessary.
Fonte: arXiv cs.LG
Privacy/Security/Fairness • Score 85
Inference-Time Machine Unlearning via Gated Activation Redirection
arXiv:2605.12765v1 Announce Type: new
Abstract: Large Language Models memorize vast amounts of training data, raising concerns regarding privacy, copyright infringement, and safety. Machine unlearning seeks to remove the influence of a targeted forget set while preserving model performance, ideally approximating a model retrained from scratch without the forget set. Existing approaches aim to achieve this by updating model parameters via gradient-based methods. However, these updates are computationally expensive, lead to irreversible weight changes, and degrade when the model is quantized for deployment. A recent alternative to changing model weights is activation engineering, where activations are changed during inference to steer model behavior. Despite circumventing weight editing, naive activation steering introduces its own failure modes, as a single global steering vector applies the same intervention to every input, leading to unintended changes in model behavior. We introduce Inference-Time Unlearning via Gated Activation Redirection (GUARD-IT), a training- and gradient-free method that unlearns via input-dependent activation steering at inference time. The resulting intervention is applied as a norm-preserving rotation in the residual stream, leaving model weights untouched. Experiments on TOFU and MUSE show that GUARD-IT matches or exceeds 12 gradient-based baselines across three model scales, while being the only method to simultaneously preserve utility, suppress memorization, and avoid catastrophic collapse across all settings. GUARD-IT further supports continual unlearning without retraining, and remains effective under quantization, a scenario in which parameter-editing methods degrade.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Identifying the nonlinear string dynamics with port-Hamiltonian neural networks
arXiv:2605.12785v1 Announce Type: new
Abstract: Hybrid machine learning combines physical knowledge with data-driven models to enhance interpretability and performance. In this context, Port-Hamiltonian Systems (PHS), which generalize Hamiltonian mechanics to describe open, non-autonomous dynamical systems, have been successfully integrated with neural networks under the name Port-Hamiltonian Neural Networks (PHNNs). While the ability of PHNNs to identify Hamiltonian ordinary differential equation (ODE) systems has already been demonstrated, their application to learning Hamiltonian partial differential equation (PDE) systems remains largely unexplored. This limitation restricts their use in musical acoustics, where instruments are typically modeled as distributed parameter systems governed by PDEs. In this work, we demonstrate how to learn the nonlinear string dynamics from data in a physically-consistent framework through a PHNN extension to PDEs. By constructing structured neural network architectures based on PHS, we can recover both the Hamiltonian governing the string and the dissipation affecting it. This approach outperforms baseline, non-physics-informed methods in terms of both accuracy and interpretability. Numerical experiments using synthetic data demonstrate the ability of the proposed PHNN model to identify and emulate the nonlinear dynamics of the system.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Ada-MK: Adaptive MegaKernel Optimization via Automated DAG-based Search for LLM Inference
arXiv:2605.11581v1 Announce Type: new
Abstract: When large language models (LLMs) serve real-time inference in commercial online advertising systems, end-to-end latency must be strictly bounded to the millisecond range. Yet every token generated during the decode phase triggers thousands of kernel launches, and kernel launch overhead alone can account for 14.6% of end-to-end inference time. MegaKernel eliminates launch overhead and inter-operator HBM round-trips by fusing multiple operators into a single persistent kernel. However, existing MegaKernel implementations face a fundamental tension between portability and efficiency on resource-constrained GPUs such as NVIDIA Ada: hand-tuned solutions are tightly coupled to specific architectures and lack portability, while auto-compiled approaches introduce runtime dynamic scheduling whose branch penalties are unacceptable in latency-critical settings. We observe that under a fixed deployment configuration, the optimal execution path of a MegaKernel is uniquely determined, and runtime dynamic decision-making can be entirely hoisted to compile time. Building on this insight, we propose Ada-MK: (1) a three-dimensional shared-memory constraint model combined with K-dimension splitting that reduces peak shared memory usage by 50%; (2) MLIR-based fine-grained DAG offline search that solidifies the optimal execution path, completely eliminating runtime branching; and (3) a heterogeneous hybrid inference engine that embeds MegaKernel as a plugin into TensorRT-LLM, combining high-throughput Prefill with low-latency Decode. On an NVIDIA L20, Ada-MK improves single-batch throughput by up to 23.6% over vanilla TensorRT-LLM and 50.2% over vLLM, achieving positive gains across all tested scenarios--the first industrial deployment of MegaKernel in a commercial online advertising system.
Fonte: arXiv cs.CL
Theory/Optimization • Score 85
Sensor Design for Accuracy-Bounded Estimation via Maximum-Entropy Likelihood Synthesis
arXiv:2605.11120v1 Announce Type: cross
Abstract: Designing the sensing architecture for large-scale spatio-temporal systems is hard when accuracy requirements are specified but sensor models are uncertain or unavailable. Classical design treats sensor placement and estimation sequentially, requiring valid forward models for each sensing modality. This paper inverts the design flow: given an error budget, synthesize the measurement likelihood that enforces it while injecting minimal information beyond the dynamical prior. The likelihood is constructed by constrained optimization: among all posteriors satisfying a prescribed accuracy bound relative to a target, select the one minimizing Kullback-Leibler divergence from the prior.
The solution is a maximum-entropy posterior in relative-entropy form, and the induced likelihood is the Radon-Nikodym derivative. The framework accommodates arbitrary discrepancies and is instantiated for Wasserstein distance, maximum mean discrepancy, $f$-divergences, moment constraints, and hybrid metrics. For each, we derive the discrete particle-level problem, analyze its convex or convex-relaxed structure, and present solvers with complexity scaling. A closed-form solution exists for the symmetric exponential-tilt case, and a distillation procedure converts nonparametric likelihood samples into parametric forms. A two-layer sensor design architecture embeds the synthesized likelihood in the recursive predict-update loop, connecting accuracy budgets to physical sensor placement, precision, and configuration. Numerical experiments comparing four metrics on unimodal and multimodal scenarios confirm the accuracy constraints are reliably enforced and reveal how metric choice determines the amount and spatial distribution of injected information.
Fonte: arXiv stat.ML
RL • Score 85
The DAWN of World-Action Interactive Models
arXiv:2605.11550v1 Announce Type: new
Abstract: A plausible scene evolution depends on the maneuver being considered, while a good maneuver depends on how the scene may evolve. Existing World Action Models (WAMs) largely miss this reciprocity, treating world prediction and action generation as either isolated parallel branches or rigid predict-then-plan pipelines. We formalize this perspective as World-Action Interactive Models (WAIMs), and instantiate it in autonomous driving with \textbf{DAWN} (\textbf{D}enoising \textbf{A}ctions and \textbf{W}orld i\textbf{N}teractive model), a simple yet strong latent generative baseline. DAWN operates in a compact semantic latent space and couples a \emph{World Predictor} with a \emph{World-Conditioned Action Denoiser}: the predicted world hypothesis conditions action denoising, while the denoised action hypothesis is fed back to update the world prediction, so that both are recursively refined during inference. Rather than eliminating test-time world evolution altogether or rolling out the full future in pixel space, DAWN performs a short explicit latent rollout that is sufficient to support long-horizon trajectory generation in complex interactive scenes. Experiments show that DAWN achieves strong planning performance and favorable safety-related results across multiple autonomous driving benchmarks. More broadly, our results suggest that interactive world-action generation is a principled path toward truly actionable world models.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Seeing the Needle in the Haystack: Towards Weakly-Supervised Log Instance Anomaly Localization via Counterfactual Perturbation
arXiv:2605.10988v1 Announce Type: new
Abstract: Log anomaly detection is a critical task for system operations and security assurance. However, in networked systems at scale, log data are generated at massive scale while instance-level annotations are prohibitively expensive, posing great difficulties to fine-grained anomaly localization. To address this challenge, we propose LogMILP (Log anomaly localization based on Multi-Instance Learning enhanced by prototypes and Perturbation), a weakly supervised framework that enables both bag-level anomaly detection and instance-level anomaly localization using only bag-level labels. Our method guides the model to pinpoint the critical log entries using prototype-guided structural modeling with counterfactual perturbation consistency regularization, thereby improving localization reliability and interpretability under coarse-grained supervision. Experimental results on three public datasets demonstrate that LogMILP achieves competitive detection performance while yielding significantly more reliable instance-level localization. Our code is open-sourced at https://github.com/YUK1207/LogMILP.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Efficient LLM-based Advertising via Model Compression and Parallel Verification
arXiv:2605.11582v1 Announce Type: new
Abstract: Large language models (LLMs) have shown remarkable potential in advertising scenarios such as ad creative generation and targeted advertising. However, deploying LLMs in real-time advertising systems poses significant challenges due to their high inference latency and computational cost. In this paper, we propose an Efficient Generative Targeting framework that integrates adaptive group quantization, layer-adaptive hierarchical sparsification, and prefix-tree parallel verification to accelerate LLM inference while preserving generation quality. Extensive experiments on two real-world advertising scenarios demonstrate that our framework achieves significant speedup with acceptable quality degradation, making it operationally viable for practical deployments.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
Pion: A Spectrum-Preserving Optimizer via Orthogonal Equivalence Transformation
arXiv:2605.12492v1 Announce Type: cross
Abstract: We introduce Pion, a spectrum-preserving optimizer for large language model (LLM) training based on orthogonal equivalence transformation. Unlike additive optimizers such as Adam and Muon, Pion updates each weight matrix through left and right orthogonal transformations, preserving its singular values throughout training. This yields an optimization mechanism that modulates the geometry of weight matrices while keeping their spectral norm fixed. We derive the Pion update rule, systematically examine its design choices, and analyze its convergence behavior along with several key properties. Empirical results show that Pion offers a stable and competitive alternative to standard optimizers for both LLM pretraining and finetuning.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
LEAP: Unlocking dLLM Parallelism via Lookahead Early-Convergence Token Detection
arXiv:2605.10980v1 Announce Type: new
Abstract: Diffusion Language Models (dLLMs) have garnered significant attention for their potential in highly parallel processing. The parallel capabilities of existing dLLMs stem from the assumption of conditional independence at high confidence levels, which ensures negligible discrepancy between the marginal and joint distributions. However, the stringent confidence thresholds required to preserve accuracy severely constrain the scalability of parallelism. Through systematic token-level statistical analysis, we reveal that a substantial proportion of tokens converge to their correct predictions early in the denoising process yet fail to reach standard confidence thresholds, confirming that current confidence-based criteria are overly conservative. In response, we introduce LEAP (Lookahead Early-Convergence Token Detection for Accelerated Parallel Decoding). LEAP is a training-free, plug-and-play method that leverages future context filtering and multi-sequence superposition to detect early-converging tokens. By validating the alignment between early convergence and correctness, we enable reliable early decoding of these tokens. Benchmarking across diverse domains demonstrates that LEAP significantly lowers inference latency and decoding steps. Compared to confidence-based decoding, the average number of denoising steps is reduced by about 30%. On the GSM8K dataset, combining LEAP with dParallel accelerates decoding to 7.2 tokens per step while preserving model precision. LEAP effectively breaks the reliance on high-confidence priors, offering a novel paradigm for parallel decoding.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Approximation Theory of Laplacian-Based Neural Operators for Reaction-Diffusion System
arXiv:2605.12025v1 Announce Type: cross
Abstract: Neural operators provide a framework for learning solution operators of partial differential equations (PDEs), enabling efficient surrogate modeling for complex systems. While universal approximation results are now well understood, approximation analysis specific to nonlinear reaction-diffusion systems remains limited. In this paper, we study neural operators applied to the solution mapping from initial conditions to time-dependent solutions of a generalized Gierer-Meinhardt reaction-diffusion system, a prototypical model of nonlinear pattern formation. Our main results establish explicit approximation error bounds in terms of network depth, width, and spectral rank by exploiting the Laplacian spectral representation of the Green's function underlying the PDE. We show that the required parameter complexity grows at most polynomially with respect to the target accuracy, demonstrating that Laplacian eigenfunction-based neural operator architectures alleviate the curse of parametric complexity encountered in generic operator learning. Numerical experiments on the Gierer-Meinhardt system support the theoretical findings.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Training-Inference Consistent Segmented Execution for Long-Context LLMs
arXiv:2605.11744v1 Announce Type: new
Abstract: Transformer-based large language models face severe scalability challenges in long-context generation due to the computational and memory costs of full-context attention. Under practical computation and memory constraints, many inference-efficient long-context methods improve efficiency by adopting bounded-context or segment-level execution only during inference, while continuing to train models under full-context attention, resulting in a mismatch between training and inference execution and state-transition semantics. Based on this insight, we propose a training-inference consistent segment-level generation framework, in which training and inference follow the same segment-level forward execution semantics. During training, consistency with inference is enforced by restricting gradient propagation to KV states carried over from the immediately preceding segment, while permitting head-specific access to past KV states during the forward pass without involving them in gradient propagation. Across long-context benchmarks, our approach achieves performance comparable to full-context attention, while achieving competitive latency-memory trade-offs against strong inference-efficient baselines, and substantially improving scalability at very long context lengths (e.g., approximately 6x lower peak prefill memory at 128K compared to full-context attention with FlashAttention).
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
COSMOS: Model-Agnostic Personalized Federated Learning with Clustered Server Models and Pseudo-Label-Only Communication
arXiv:2605.11165v1 Announce Type: new
Abstract: Federated learning (FL) in heterogeneous environments remains challenging because client models often differ in both architecture and data distribution. While recent approaches attempt to address this challenge through client clustering and knowledge distillation, simultaneously handling architectural and statistical heterogeneity remains difficult. We introduce COSMOS, a model-agnostic framework that enables server-side personalization using only pseudo-label communication. Clients train local models and predict on the public data; the server clusters clients by prediction similarity, trains a cluster-specific model for each group using its own compute, and distills the resulting models back to clients. We provide the first theoretical analysis showing that distillation from the learned cluster models can yield exponential personalization risk contraction, going beyond the convergence-to-stationarity guarantees typically provided in model-agnostic FL. Experiments across benchmarks demonstrate that COSMOS consistently outperforms all model-agnostic FL baselines while remaining competitive with state-of-the-art personalized FL methods. More broadly, our results highlight personalized server-side learning with pseudo-labels as a promising paradigm for scalable and model-agnostic federated learning in highly heterogeneous environments.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
FibQuant: Universal Vector Quantization for Random-Access KV-Cache Compression
arXiv:2605.11478v1 Announce Type: cross
Abstract: Long-context inference is increasingly a memory-traffic problem. The culprit is the key--value (KV) cache: it grows with context length, batch size, layers, and heads, and it is read at every decoding step. Rotation-based scalar codecs meet this systems constraint by storing a norm, applying a shared random rotation, and quantizing one coordinate at a time. They are universal and random-access, but they discard the geometry created by the normalization step. After a Haar rotation, a block of $k$ consecutive coordinates is not a product source; it is a spherical-Beta source on the unit ball. We introduce \textsc{FibQuant}, a universal fixed-rate vector quantizer that keeps the same normalize--rotate--store interface while replacing scalar tables by a shared radial--angular codebook matched to this canonical source. The codebook combines Beta-quantile radii, Fibonacci\,/\,Roberts--Kronecker quasi-uniform directions, and multi-restart Lloyd--Max refinement. We prove that the resulting vector code strictly improves on its scalar product specialization at matched rate, with a high-rate gain that separates into a cell-shaping factor and a density-matching factor. The same construction gives a dense rate axis, including fractional-bit and sub-one-bit operating points, without calibration or variable-length addresses. On GPT-2 small KV caches, \textsc{FibQuant} traces a memory--fidelity frontier from $5\times$ compression at $0.99$ attention cosine similarity to $34\times$ at $0.95$. End-to-end on TinyLlama-1.1B, it is within $0.10$ perplexity of fp16 at $4\times$ compression and has $3.6\times$ lower perplexity than scalar \textsc{TurboQuant} at $b = 2$ ($8\times$ compression), where scalar random-access quantization begins to fail.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Rotation-Preserving Supervised Fine-Tuning
arXiv:2605.10973v1 Announce Type: new
Abstract: Supervised fine-tuning (SFT) improves in-domain performance but can degrade out-of-domain (OOD) generalization. Prior work suggests that this degradation is related to changes in dominant singular subspaces of pretrained weight matrices. However, directly identifying loss-sensitive directions with Hessian or Fisher information is computationally expensive at LLM scale. In this work, we propose preserving projected rotations in pretrained singular subspaces as an efficient proxy for Fisher-sensitive directions, which we call Rotation-Preserving Supervised Fine-Tuning (RPSFT). RPSFT penalizes changes in the projected top-$k$ singular-vector block of each pretrained weight matrix, limiting unnecessary rotation while preserving task adaptation. Across model families and sizes trained on math reasoning data, RPSFT improves the in-domain/OOD trade-off over standard SFT and strong SFT baselines, better preserves pretrained representations, and provides stronger initializations for downstream RL fine-tuning. Code is available at \href{https://github.com/jinhangzhan/RPSFT.git}{https://github.com/jinhangzhan/RPSFT}.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
A Comparative Study of Federated Learning Aggregation Strategies under Homogeneous and Heterogeneous Data Distributions
arXiv:2605.11010v1 Announce Type: new
Abstract: Federated Learning has emerged as a transformative paradigm for collaborative machine learning across distributed environments. However, its performance is strongly influenced by the aggregation strategy used to combine local model updates at the server, which directly affects learning performance, robustness, and system behavior. This work presents a comprehensive experimental comparison of widely used federated aggregation strategies under both homogeneous and heterogeneous data distributions. Using benchmark image classification datasets, we analyze how different aggregation mechanisms respond to varying degrees of data heterogeneity, examining their impact on centralized accuracy and loss, and system-level efficiency metrics, including aggregation, training, and communication time. The results demonstrate that aggregation strategies exhibit distinct trade-offs across datasets and data distributions, with their effectiveness varying according to dataset characteristics and operating conditions.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
DisagMoE: Computation-Communication overlapped MoE Training via Disaggregated AF-Pipe Parallelism
arXiv:2605.11005v1 Announce Type: new
Abstract: Mixture-of-experts (MoE) architectures enable trillion-parameter LLMs with sparsely activated experts. Expert parallelism (EP) is a widely adopted MoE training strategy, but it suffers from severe all-to-all communication bottlenecks, which is exaggerated by the limited inter-node network bandwidth as the growing model size requires distributing experts across GPU nodes. Prior work focused on overlapping these all-to-all communications with feed-forward network (FFN) and self-attention computations, which often leaves residual network-bound stalls due to inherent imbalance in attention and FFN layers' computation-communication ratios. We present DisagMoE, a disaggregated MoE training system that jointly optimizes model placement and scheduling for maximal efficiency. DisagMoE separates attention and FFN layers into disjoint GPU groups, introduces a multi-stage pipeline with uni-directional, many-to-many communications, and employs a computation-communication roofline model to balance GPU and network bandwidth allocation among the attention and FFN groups. DisagMoE is implemented on Megatron-LM, and evaluation shows that DisagMoE improves training efficiency across multiple MoE models with up to 1.8x speedup on 16-node 8xH800 clusters.
Fonte: arXiv cs.LG
Privacy/Security/Fairness • Score 85
Robust LLM Unlearning Against Relearning Attacks: The Minor Components in Representations Matter
arXiv:2605.11685v1 Announce Type: new
Abstract: Large language model (LLM) unlearning aims to remove specific data influences from pre-trained model without costly retraining, addressing privacy, copyright, and safety concerns. However, recent studies reveal a critical vulnerability: unlearned models rapidly recover "forgotten" knowledge through relearning attacks. This fragility raises serious security concerns, especially for open-weight models. In this work, we investigate the fundamental mechanism underlying this fragility from a representation geometry perspective. We discover that existing unlearning methods predominantly optimize along dominant components, leaving minor components largely unchanged. Critically, during relearning attacks, the modifications in these dominant components are easily reversed, enabling rapid knowledge recovery, whereas minor components exhibit stronger resistance to such reversal. We further provide a theoretical analysis that explains both observations from the spectral structure of representations. Building on this insight, we propose Minor Component Unlearning (MCU), a novel unlearning approach that explicitly targets minor components in representations. By concentrating unlearning effects in these inherently robust directions, our method achieves substantially improved resistance to relearning attacks. Extensive experiments on three datasets validate our approach, demonstrating significant improvements over state-of-the-art methods including sharpness-aware minimization.
Fonte: arXiv cs.CL
MLOps/Systems • Score 85
PD-4DGS:Progressive Decomposition of 4D Gaussian Splatting for Bandwidth-Adaptive Dynamic Scene Streaming
arXiv:2605.11427v1 Announce Type: new
Abstract: 4D Gaussian Splatting (4DGS) enables high-quality dynamic novel view synthesis, yet current models remain monolithic bitstreams that clients must download in full before any frame can be rendered, causing black-screen waits of tens to hundreds of seconds on mobile bandwidth and leaving 4DGS incompatible with modern adaptive-bitrate delivery. Progressive 3DGS compression alleviates this for static scenes, but it acts only on spatial anchors and cannot partition the temporal deformation networks that dominate dynamic-scene size. We present PD-4DGS, the first framework for progressive compression and on-demand transmission of 4DGS. Hierarchical Deformation Decomposition (HDD) externalises the coarse-to-fine motion hierarchy already latent in 4DGS into three independently transmittable layers -- a static scaffold, a global deformation, and a local refinement -- so that any prefix of the bitstream is already renderable, turning a single training run into a scalable, DASH/HLS-compatible bitstream. A Gaussian-entropy attribute rate-distortion loss together with a temporal mask consistency regulariser shrink the base layer while suppressing low-bitrate flicker; a capacity-weighted rollout schedule, gated online by a learnt activation rate rho, then prevents deformation-network under-training without any per-scene hyperparameter. On the Dycheck iPhone benchmark, PD-4DGS cuts the streamed bitstream by >60% at matched rendering fidelity and reduces first-frame latency from 73--930 s to ~1.7 s on a 2 Mbps link, uniquely enabling true on-demand progressive streaming for 4DGS.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
Enabling Performant and Flexible Model-Internal Observability for LLM Inference
arXiv:2605.11093v1 Announce Type: new
Abstract: Today's inference-time workloads increasingly depend on timely access to a model's internal states. We present DMI-Lib, a high-speed deep model inspector that treats internal observability as a first-class systems primitive, decoupling it from the inference hot path via an asynchronous observability substrate built from Ring^2, a GPU-CPU memory abstraction for capturing and staging tensors, and a policy-controlled host backend that exports them. DMI-Lib enables the placement of observation points across a rich space of internal signals and diverse inference backends while preserving serving optimizations and adhering to tight GPU memory budgets. Our experiments demonstrate that DMI-Lib incurs only 0.4%--6.8% overhead in offline batch inference and an average of 6% in moderate online serving, reducing latency overhead by 2x-15x compared to existing baselines with similar observability features. DMI-Lib is open-sourced at https://github.com/ProjectDMX/DMI.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
LOFT: Low-Rank Orthogonal Fine-Tuning via Task-Aware Support Selection
arXiv:2605.11872v1 Announce Type: cross
Abstract: Orthogonal parameter-efficient fine-tuning (PEFT) adapts pretrained weights through structure-preserving multiplicative transformations, but existing methods often conflate two distinct design choices: the subspace in which adaptation occurs and the transformation applied within that subspace. This paper introduces LOFT, a low-rank orthogonal fine-tuning framework that explicitly separates these two components. By viewing orthogonal adaptation as a multiplicative subspace rotation, LOFT provides a unified formulation that recovers representative orthogonal PEFT methods, including coordinate-, butterfly-, Householder-, and principal-subspace-based variants. More importantly, this perspective exposes support selection as a central design axis rather than a byproduct of a particular parameterization. We develop a first-order analysis showing that useful adaptation supports should be informed by the downstream training signal, motivating practical task-aware support selection strategies. Across language understanding, visual transfer, mathematical reasoning, and multilingual out-of-distribution adaptation, LOFT recovers principal-subspace orthogonal adaptation while gradient-informed supports improve the efficiency-performance trade-off under matched parameter, memory, and compute budgets. These results suggest that principled support selection is an important direction for improving orthogonal PEFT.
Fonte: arXiv stat.ML
Multimodal • Score 85
OmniThoughtVis: A Scalable Distillation Pipeline for Deployable Multimodal Reasoning Models
arXiv:2605.11629v1 Announce Type: new
Abstract: Recent multimodal large language models (MLLMs) have shown strong chain-of-thought (CoT) reasoning ability on vision-language tasks, but their direct deployment in real-world systems is often limited by latency and resource constraints. In practice, smaller MLLMs are preferred for online serving, yet their reasoning performance is bottlenecked by the lack of large-scale, high-quality multimodal CoT supervision. In this paper, we present OmniThoughtVis, a scalable data curation and distillation pipeline for transferring multimodal reasoning capabilities from high-capacity teacher models to smaller, deployment-oriented MLLMs. Starting from a diverse open-source seed pool, our pipeline generates structured CoT traces and performs joint annotation of reasoning difficulty, answer quality, and semantic task tags. To maintain data quality at scale, we combine rule-based filtering, difficulty-aware selection, and tag-based diversity sampling, resulting in a curated corpus of 1.8M samples that supports controllable subset construction for downstream training. We use OmniThoughtVis to distill Qwen3-VL models from 2B to 8B parameters and evaluate them on nine multimodal reasoning benchmarks. The resulting distilled models show consistent gains across model scales, including improvements of up to +16.8 points on MathVerse and +5.6 points on MMMU-Pro for the 4B model. Notably, the distilled 4B model matches or surpasses the undistilled 8B baseline on several tasks, highlighting the practical value of scalable reasoning distillation for deployment-oriented MLLMs.
Fonte: arXiv cs.CL
Vision • Score 85
LatentHDR: Decoupling Exposure from Diffusion via Conditional Latent-to-Latent Mapping for Text/Image-to-Panoramic HDR
arXiv:2605.11115v1 Announce Type: new
Abstract: High Dynamic Range (HDR) generation remains challenging for generative models, which are largely limited to low dynamic range outputs. Recent diffusionbased approaches approximate HDR by generating multiple exposure-conditioned samples, incurring high computational cost and structural inconsistencies across exposures. We propose LatentHDR, a framework that decouples scene generation from exposure modeling in latent space. A pretrained diffusion backbone produces a single coherent scene representation, while a lightweight conditional latent to-latent head deterministically maps it to exposure-specific representations. This enables the generation of a dense, structurally consistent exposure stack in a single pass. This design eliminates multi-pass diffusion, ensures cross-exposure alignment, and enables scalable HDR synthesis. LatentHDR supports both textand image-conditioned HDR generation for perspective and panoramic scenes. Experiments on synthetic data and the SI-HDR benchmark show that LatentHDR achieves state-of-the-art dynamic range with competitive perceptual quality, while reducing computation by an order of magnitude. Our results demonstrate that high-quality HDR generation can be achieved through structured latent modeling, challenging the need for stochastic multi-exposure generation.
Fonte: arXiv cs.CV
Vision • Score 85
Lite3R: A Model-Agnostic Framework for Efficient Feed-Forward 3D Reconstruction
arXiv:2605.11354v1 Announce Type: new
Abstract: Transformer-based 3D reconstruction has emerged as a powerful paradigm for recovering geometry and appearance from multi-view observations, offering strong performance across challenging visual conditions. As these models scale to larger backbones and higher-resolution inputs, improving their efficiency becomes increasingly important for practical deployment. However, modern 3D transformer pipelines face two coupled challenges: dense multi-view attention creates substantial token-mixing overhead, and low-precision execution can destabilize geometry-sensitive representations and degrade depth, pose, and 3D consistency. To address the first challenge, we propose Lite3R, a model-agnostic teacher-student framework that replaces dense attention with Sparse Linear Attention to preserve important geometric interactions while reducing attention cost. To address the second challenge, we introduce a parameter-efficient FP8-aware quantization-aware training (FP8-aware QAT) strategy with partial attention distillation, which freezes the vast majority of pretrained backbone parameters and trains only lightweight linear-branch projection layers, enabling stable low-precision deployment while retaining pretrained geometric priors. We further evaluate Lite3R on two representative backbones, VGGT and DA3-Large, over BlendedMVS and DTU64, showing that it substantially reduces latency (1.7-2.0x) and memory usage (1.9-2.4x) while preserving competitive reconstruction quality overall. These results demonstrate that Lite3R provides an effective algorithm-system co-design approach for practical transformer-based 3D reconstruction. Code: https://github.com/AIGeeksGroup/Lite3R. Website: https://aigeeksgroup.github.io/Lite3R.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
SkillGen: Verified Inference-Time Agent Skill Synthesis
arXiv:2605.10999v1 Announce Type: new
Abstract: Skills are a promising way to improve LLM agent capabilities without retraining, while keeping the added procedure reusable and controllable. However, high-quality skills are still largely written by hand. We introduce SkillGen, a multi-agent framework that synthesizes a single auditable skill from trajectories generated by a base agent. The output is a human-readable artifact that can be inspected before use. Rather than merely summarizing trajectories, SkillGen leverages contrastive induction over both successful and failed trajectories to identify reusable success patterns, recurring failure modes, and behaviors that appear in nearby successes but are missing from failures. SkillGen then generates candidate skills and iteratively refines the skill. A key novelty in SkillGen is that we model agent skills as interventions to empirically verify the net effect of skills on the overall performance. Specifically, we compare outcomes on the same instances with and without the skill, so that we account for both repairs (cases where the skill fixes a baseline failure) and regressions (cases where the skill breaks a baseline success). Across a broad range of agents and datasets, SkillGen consistently improves held-out performance, outperforms existing skill-generation baselines, and produces skills that transfer across models.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
CATS: Cascaded Adaptive Tree Speculation for Memory-Limited LLM Inference Acceleration
arXiv:2605.11186v1 Announce Type: new
Abstract: Auto-regressive decoding in Large Language Models (LLMs) is inherently memory-bound: every generation step requires loading the model weights and intermediate results from memory (e.g., High-Bandwidth Memory (HBM) for GPU servers), making throughput bottlenecked by memory bandwidth rather than compute. Speculative decoding addresses this by enabling parallel verification of multiple draft tokens, effectively amortizing the cost of each target-model call. However, existing speculative decoding methods are designed under the assumption that HBM is sufficiently large to hold both the target model and an auxiliary draft model simultaneously -- an assumption that breaks down on memory-constrained devices such as edge platforms with limited DRAM. We analyze the inference bottleneck in this memory-limited regime and propose CATS, a self-speculative decoding framework that conducts cascaded verification and correction based on the memory budget and parameter offloading patterns on memory-limited devices. This design maximizes token acceptance rate and end-to-end speedup while keeping the peak memory footprint on the device equal to that of the target model alone. We evaluate CATS on different models across five benchmarks on real edge devices. CATS can achieve a wall-clock speedup of up to 5.08x with no degradation in generation quality, outperforming the SOTA method by up to 1.45x under edge memory constraints.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
SOMA: Efficient Multi-turn LLM Serving via Small Language Model
arXiv:2605.11317v1 Announce Type: new
Abstract: Large Language Models (LLMs) are increasingly deployed in multi-turn dialogue settings where preserving conversational context across turns is essential. A standard serving practice concatenates the full dialogue history at every turn, which reliably maintains coherence but incurs substantial cost in latency, memory, and API expenditure, especially when queries are routed to large proprietary models. Existing approaches often struggle to balance the trade-off between response quality and efficiency. We propose a framework that exploits the early turns of a session to estimate a local response manifold and then adapt a smaller surrogate model to this local region for the remainder of the conversation. Concretely, we learn soft prompts that maximize semantic divergence between the large and surrogate small language models' responses to surface least-aligned local directions, stabilize training with anti-degeneration control, and distill the mined cases into localized LoRA fine-tuning so the surrogate runs without prompts at inference. A simple gate enables a one-time switch with rollback on drift. We further provide a theoretical analysis for key components in SOMA. Extensive experiments show the effectiveness of SOMA. The source code is provided at: https://github.com/LabRAI/SOMA.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
Variational Linear Attention: Stable Associative Memory for Long-Context Transformers
arXiv:2605.11196v1 Announce Type: new
Abstract: Linear attention reduces the quadratic cost of softmax attention to $\mathcal{O}(T)$, but its memory state grows as $\mathcal{O}(T)$ in Frobenius norm, causing progressive interference between stored associations. We introduce \textbf{Variational Linear Attention} (VLA), which reframes the memory update as an online regularised least-squares problem with an adaptive penalty matrix maintained via the Sherman-Morrison rank-1 formula. We prove that normalising the write direction to unit length gives the recurrence Jacobian spectral norm exactly $1$ for all sequence lengths and head dimensions (Proposition 2), and that the state norm is self-limiting under bounded inputs (Proposition 1). Empirically, VLA reduces $\|S_t\|_F$ by $109\times$ relative to standard linear attention at $T{=}1{,}000$, achieves near-perfect exact-match accuracy on multi-query associative recall within the effective per-head memory regime ($n_\text{pairs} < d_h$), maintaining substantially higher retrieval performance than DeltaNet and standard linear attention under increasing memory load, and maintains 62\% accuracy at the per-head capacity boundary. A Triton-fused kernel achieves $14\times$ speedup over sequential Python and $\mathcal{O}(T)$ scaling, crossing below softmax attention latency at approximately 43\,000 tokens.
Fonte: arXiv cs.LG
Privacy/Security/Fairness • Score 85
Unlearning with Asymmetric Sources: Improved Unlearning-Utility Trade-off with Public Data
arXiv:2605.11170v1 Announce Type: new
Abstract: Noise-based certified machine unlearning currently faces a hard ceiling: the noise magnitude required to certify unlearning typically destroys model utility, particularly for large-scale deletion requests.
While leveraging public data is a standard technique in differential privacy to relax this tension, its role in unlearning remains unexplored. We address this gap by introducing Asymmetric Langevin Unlearning (ALU), a framework that uses public data to mitigate privacy costs. We prove that public data injection suppresses the unlearning cost by a factor of $O(1/n_{\mathrm{pub}}^2)$, guaranteeing a strict computational advantage over retraining. This establishes a new control mechanism: practitioners can mitigate the need for high noise-and the associated utility loss-by increasing the volume of public data. Crucially, we analyze the realistic setting of distribution mismatch, explicitly characterizing how shifts between public and private sources impact utility.
We show that ALU enables mass unlearning of constant dataset fractions -- a regime where standard symmetric methods become impractical -- while maintaining high utility. Empirical evaluations using variational R\'enyi divergence and membership inference attacks confirm that ALU effectively thwarts privacy attacks while preserving utility under reasonable distribution shifts.
Fonte: arXiv cs.LG
Vision • Score 85
USEMA: a Scalable Efficient Mamba Like Attention for Medical Image Segmentation
arXiv:2605.11131v1 Announce Type: new
Abstract: Accurate medical image segmentation is an integral part of the medical image analysis pipeline that requires the ability to merge local and global information. While vision transformers are able to capture global interactions using vanilla self-attention, their quadratic computational complexity in the input size remains a struggle for medical image segmentation tasks. Motivated by the dispersion property of vanilla self-attention and recent development of Mamba form of attention, Scalable and Efficient Mamba like Attention (SEMA) utilizes token localization via local window attention to avoid dispersion and maintain focusing, complemented by theoretically consistent arithmetic averaging to capture global aspect of attention. In this work, we present USEMA, a hybrid UNet architecture that merges the local feature extraction ability of convolutional neural networks (CNNs) with SEMA attention. We conduct experiments with USEMA across a variety of modalities and image sizes, demonstrating improved computational efficiency compared to transformer based models using full self-attention, and superior segmentation performance relative to purely convolution and Mamba-based models.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Learning to Foresee: Unveiling the Unlocking Efficiency of On-Policy Distillation
arXiv:2605.11739v1 Announce Type: new
Abstract: On-policy distillation (OPD) has emerged as an efficient post-training paradigm for large language models. However, existing studies largely attribute this advantage to denser and more stable supervision, while the parameter-level mechanisms underlying OPD's efficiency remain poorly understood. In this work, we argue that OPD's efficiency stems from a form of ``foresight'': it establishes a stable update trajectory toward the final model early in training. This foresight manifests in two aspects. First, at the \textbf{Module-Allocation Level}, OPD identifies regions with low marginal utility and concentrates updates on modules that are more critical to reasoning. Second, at the \textbf{Update-Direction Level}, OPD exhibits stronger low-rank concentration, with its dominant subspaces aligning closely with the final update subspace early in training. Building on these findings, we propose \textbf{EffOPD}, a plug-and-play acceleration method that speeds up OPD by adaptively selecting an extrapolation step size and moving along the current update direction. EffOPD requires no additional trainable modules or complex hyperparameter tuning, and achieves an average training acceleration of $3\times$ while maintaining comparable final performance. Overall, our findings provide a parameter-dynamics perspective for understanding the efficiency of OPD and offer practical insights for designing more efficient post-training methods for large language models.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
Semi-Supervised Bayesian GANs with Log-Signatures for Uncertainty-Aware Credit Card Fraud Detection
arXiv:2509.00931v3 Announce Type: replace
Abstract: We present a novel deep generative semi-supervised framework for credit card fraud detection, formulated as time series classification task. As financial transaction data streams grow in scale and complexity, traditional methods often require large labeled datasets, struggle with time series of irregular sampling frequencies and varying sequence lengths. To address these challenges, we extend conditional Generative Adversarial Networks (GANs) for targeted data augmentation, integrate Bayesian inference to obtain predictive distributions and quantify uncertainty, and leverage log-signatures for robust feature encoding of transaction histories. We introduce a novel Wasserstein distance-based loss to align generated and real unlabeled samples while simultaneously maximizing classification accuracy on labeled data. Our approach is evaluated on the BankSim dataset, a widely used simulator for credit card transaction data, under varying proportions of labeled samples, demonstrating consistent improvements over benchmarks in both global statistical and domain-specific metrics. These findings highlight the effectiveness of GAN-driven semi-supervised learning with log-signatures for irregularly sampled time series and emphasize the importance of uncertainty-aware predictions.
Fonte: arXiv stat.ML
MLOps/Systems • Score 85
An Empirical Study of Automating Agent Evaluation
arXiv:2605.11378v1 Announce Type: new
Abstract: Agent evaluation requires assessing complex multi-step behaviors involving tool use and intermediate reasoning, making it costly and expertise-intensive. A natural question arises: can frontier coding assistants reliably automate this evaluation process? Our study shows that simply prompting coding assistants is insufficient for this task. Without domain-specific evaluation knowledge, frontier coding assistants achieve only a 30% execution success rate and produce over-engineered evaluations averaging 12+ metrics per agent, indicating that strong coding ability does not automatically translate to reliable agent evaluation. We introduce EvalAgent, an AI assistant that automates the end-to-end agent evaluation pipeline. EvalAgent encodes evaluation domain expertise as evaluation skills (procedural instructions, reusable code and templates, and dynamically retrieved API documentation) that compose into a trace-based pipeline producing complete evaluation artifacts including metrics, executable code, and reports. To systematically assess generated evaluations, we introduce a meta-evaluation framework alongside AgentEvalBench, a benchmark comprising 20 agents, each paired with evaluation requirements and test scenarios. We further propose the Eval@1 metric to measure whether generated evaluation code both executes and yields meaningful results on the first run. Our experiments show that EvalAgent produces focused evaluations, improving Eval@1 from 17.5% to 65%, and achieving 79.5% human expert preference over baseline approaches. Further ablation studies show that evaluation skills are critical for handling complex evaluation: removing them causes Eval@1 to drop significantly from 65% to 30%.
Fonte: arXiv cs.CL
MLOps/Systems • Score 85
AESOP: Adversarial Execution-path Selection to Overload Deep Learning Pipelines
arXiv:2605.10987v1 Announce Type: new
Abstract: Modern machine learning deployments increasingly compose specialized models into dynamic inference pipelines, where upstream components produce intermediate predictions that determine the workload and inputs of downstream components. The cost of processing an input is therefore not determined by any single model, but by two coupled factors: the per-inference cost of each invoked component and its workload volume. Because these pipelines run under hard real-time constraints, efficiency is a fundamental requirement for system availability. We show that this structure creates an efficiency-attack surface that existing methods targeting single models cannot exploit: on identical inputs and budgets, path-aware targeting inflates FLOPs by $2,407\times$ while the strongest single-model baseline achieves $117\times$ -- a $20\times$ gap attributable entirely to where the attack is directed. We formalize this as the adversarial path-selection problem and present AESOP, a framework combining vulnerability-guided path ranking with adaptive loss weighting. We evaluate AESOP on five pipelines plus a production-realistic deployment variant with batching, bounded buffering, and confidence-threshold defenses. AESOP achieves up to $2,407\times$ FLOPs and $419\times$ latency inflation in white-box setting and 58$\times$ FLOPs / 17$\times$ latency in gray-box settings. Under system-level defenses, the attack is not neutralized but redirected: pipelines are forced to choose between throughput collapse ($0.578 \to 0.006$ input/s) and $96.7\%$ data loss to sustain throughput.
Fonte: arXiv cs.LG
Multimodal • Score 85
ReVision: Scaling Computer-Use Agents via Temporal Visual Redundancy Reduction
arXiv:2605.11212v1 Announce Type: new
Abstract: Computer-use agents~(CUAs) rely on visual observations of graphical user interfaces, where each screenshot is encoded into a large number of visual tokens. As interaction trajectories grow, the token cost increases rapidly, limiting the amount of history that can be incorporated under fixed context and compute budgets. This has resulted in no or very limited improvement in the performance when using history unlike other domains. We address this inefficiency by introducing ReVision, which is used to train multimodal language models on trajectories where redundant visual patches are removed using a learned patch selector that compares patch representations across consecutive screenshots while preserving spatial structure required by the model. Across three benchmarks, OSWorld, WebTailBench, and AgentNetBench, when processing trajectories with 5 history screenshots using Qwen2.5-VL-7B, ReVision reduces token usage by approximately 46% on average while improving success rate by 3% over the no drop baseline. This establishes a clear efficiency gain, enabling agents to process longer trajectories with fewer tokens. With this improved efficiency, we revisit the role of history in CUAs and find that performance continues to improve as more past observations are incorporated when redundancy is removed. This suggests that the commonly observed saturation in visual history is not due to limited usefulness of past information, but rather a consequence of inefficient token representations.
Fonte: arXiv cs.CL
MLOps/Systems • Score 85
QuIDE: Mastering the Quantized Intelligence Trade-off via Active Optimization
arXiv:2605.10959v1 Announce Type: new
Abstract: There is currently no unified metric for evaluating the efficiency of quantized neural networks. We propose QuIDE, built around the Intelligence Index I = (C x P)/log_2(T+1), which collapses the compression-accuracy-latency trade-off into a single score. Experiments across six settings -- SimpleCNN (MNIST, CIFAR), ResNet-18 (ImageNet-1K), and Llama-3-8B -- show a task-dependent Pareto Knee. 4-bit quantization is optimal for MNIST and large LLMs, while 8-bit is the sweet spot for complex CNN tasks (ResNet-18 on ImageNet), where 4-bit PTQ collapses accuracy catastrophically. The accuracy-gated variant I' correctly flags these non-viable configurations that the raw I would reward. QuIDE provides a reproducible evaluation protocol and a ready-to-use fitness function for mixed-precision search.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Echo-LoRA: Parameter-Efficient Fine-Tuning via Cross-Layer Representation Injection
arXiv:2605.08177v1 Announce Type: new
Abstract: Parameter-efficient fine-tuning (PEFT) has become a practical route for adapting large language models to downstream tasks, with LoRA-style methods being particularly attractive because they are inexpensive to train and easy to deploy. Most LoRA variants, however, revise the update rule within the weight space of each layer and leave the intermediate representations formed by deeper layers largely unused. We propose Echo-LoRA, a cross-layer representation injection method for parameter-efficient fine-tuning. During training, Echo-LoRA collects boundary hidden states from deeper source layers, aggregates them into a sample-level echo representation, and uses lightweight projection and gating networks to inject the resulting signal into shallow LoRA or DoRA modules. Answer-only masking, masked distillation, and stochastic routing are used to keep this auxiliary path stable and to reduce the gap between training and inference. On eight commonsense reasoning benchmarks, Echo-LoRA exceeds the reported LoRA baselines by 5.7 percentage points on average across LLaMA-7B, LLaMA2-7B, and LLaMA3-8B. Under reproduced LoRA baselines in our unified implementation, the average gain is 3.0 points; when combined with DoRA, the gain is 2.7 points. The Echo path is discarded after training, so the deployed model keeps the original low-rank LoRA/DoRA form and adds neither inference-time parameters nor inference computation.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
TRAM: Training Approximate Multiplier Structures for Low-Power AI Accelerators
arXiv:2605.08231v1 Announce Type: new
Abstract: Reducing power consumption in AI accelerators is increasingly important. Approximate computing can reduce power consumption while keeping the accuracy loss small. Since multipliers are power-hungry components in AI models, this paper focuses on synthesizing low-power approximate multipliers (AxMs). Unlike prior works that design AxMs separately from AI model training, we present TRAM, which jointly optimizes the AxM structure and AI model parameters to lower power with small accuracy loss. Experiments show that compared to state-of-the-art AxMs, TRAM achieves up to 25.05% AxM power reduction on CNNs with CIFAR-10, and reduces power by up to 27.09% on vision transformers with ImageNet.
Fonte: arXiv cs.LG
Vision • Score 85
Test-Time Training for Visual Foresight Vision-Language-Action Models
arXiv:2605.08215v1 Announce Type: new
Abstract: Visual Foresight VLA (VF-VLA) has become a prominent architectural choice in the recent VLA due to its impressive performance. Nevertheless, the inherent design of VF-VLA makes it particularly vulnerable to out-of-distribution (OOD) shifts. Because the quality of action directly depends on the accuracy of the predicted future visual information, OOD conditions affect both stages at once. To address this vulnerability, we propose Test-Time Training Visual Foresight VLA ($T^3$VF), a test-time training approach motivated by the observation that the predicted future image and its subsequent observation form a natural supervision pair. To further address the practical challenges that arise from indiscriminate test-time updates, we introduce an adaptive update filtering mechanism. Empirically, $T^3$VF mitigates the OOD vulnerability of VF-VLA at a modest additional inference cost, without requiring any architectural modification or auxiliary modules.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
Synergistic Simplex: Cooperative Runtime Assurance for Safety-Critical Autonomous Systems
arXiv:2605.08190v1 Announce Type: new
Abstract: Autonomous systems increasingly rely on machine-learning (ML) components for safety-critical tasks such as perception and control in autonomous vehicles (AVs). While ML enables essential capabilities, it inevitably exhibits long-tail faults that make it unsuitable for safety-critical tasks. Runtime assurance (RTA) mitigates this issue by pairing ML components with verifiable safety monitors, e.g., Control Simplex and Perception Simplex architectures. However, the limited performance of safety monitors remains a major bottleneck.
The Synergistic Simplex (SS) architecture improves system performance by enabling bidirectional integration between ML components and safety monitors while preserving formal safety guarantees. The key innovation here is allowing safety monitors to use ML outputs, which is typically prohibited in RTA systems. We formally derive conditions under which this integration preserves safety and demonstrate the performance benefits. We present the design, analysis, and evaluation of SS for AV obstacle detection.
Fonte: arXiv cs.LG
Vision • Score 85
PaceVGGT: Pre-Alternating-Attention Token Pruning for Visual Geometry Transformers
arXiv:2605.08371v1 Announce Type: new
Abstract: Visual Geometry Transformer (VGGT) is a strong feed-forward model for multiple 3D tasks, but its Alternating-Attention (AA) stack scales quadratically in the total token count, making long clips expensive.
Existing token-reduction accelerators operate inside AA, leaving the patch grid that enters AA uncompressed.
We introduce PaceVGGT, a pre-AA token pruning framework that prunes DINO patch tokens before the first AA block of a frozen VGGT. PaceVGGT trains a lightweight Token Scorer that estimates per-token importance from DINO features. The scorer is first distilled against an AA-internal attention target from the unpruned backbone, then refined under downstream camera, depth, and point-map losses.
A per-frame keep budget fixes the backbone-visible sequence length, while an importance-adaptive merge/prune assignment preserves residual content from high-saliency frames under a fixed total merge budget.
A Feature-guided Restoration module reconstructs the dense spatial grid required by the prediction heads.
On ScanNet-50 and 7-Scenes, PaceVGGT remains on the reconstruction quality--latency frontier while reducing inference latency.
On ScanNet-50, it reduces latency by \(5.1\times\) over unmodified VGGT at \(N=300\) and \(1.47\times\) over LiteVGGT at \(N=1000\).
These results identify pre-AA pruning as a viable acceleration route for frozen VGGT-style geometry transformers.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
Delivering Science as a Service: Sci-Orchestra's Cloud-Native Approach to HPC
arXiv:2605.08396v1 Announce Type: new
Abstract: The increasing complexity of modern computational environments often burdens researchers with infrastructure management, authentication protocols, and container deployments. We present Sci-Orchestra, a layered orchestration framework designed to fully automate experimental workflows, allowing scientists to prioritize scientific discovery over backend operations. By abstracting execution through an API-driven interface, the system assumes responsibility for secure authentication, resource management, and scalable deployment across diverse high-performance computing environments using Kubernetes architectures. A key innovation of Sci-Orchestra is its autonomous marketplace, which serves as a catalyst for cross-institutional collaboration. Through an intuitive user interface, researchers can rapidly deploy and share specialized services via simple selections, eliminating the need for complex installations and technical setups. This modular infrastructure is specifically designed to facilitate industry partnerships as it provides a secure execution environment and allows external collaborators to test and validate proprietary tools without the need for source-code exchange. This ``black-box'' interoperability protects intellectual property while enabling seamless integration into broader scientific pipelines, ultimately accelerating the transition from laboratory prototypes to industrial-scale applications.
Fonte: arXiv cs.CV
Vision • Score 85
Advanced Tumor Segmentation in PET/CT Imaging: A Training Strategy Study with nnU-Net for AutoPET III
arXiv:2605.08161v1 Announce Type: new
Abstract: Tumor segmentation in whole-body PET/CT imaging is crucial for precise disease evaluation and treatment planning. However, it remains challenging due to variability in lesion size, contrast, and anatomical distribution. Relying on manual segmentation makes the process time-consuming and prone to intra- and inter-observer variability. This work presents a whole-body tumor segmentation method developed for the AutoPET III challenge, where the goal is to build models that generalize across tracers and multi-center data. We employ the nnU-Net framework with a ResNet-based encoder as our baseline and systematically investigate the impact of training strategies, including intensity normalization, batch dice optimization, and data augmentation using CraveMix. Our experiments show that these strategies significantly influence model performance, particularly in reducing false positives and improving robustness to lesion variability. The best-performing configuration achieves a Dice score of up to 0.80 on the preliminary test phase, and our method ranked third in the AutoPET III challenge. The code is publicly available here.
Fonte: arXiv cs.CV
Theory/Optimization • Score 85
Posterior Concentration of Bayesian Physics-Informed Neural Networks for Elliptic PDEs
arXiv:2605.08672v1 Announce Type: cross
Abstract: We study the posterior contraction rate of Bayesian Physics-Informed Neural Networks (PINNs) for solving a general class of elliptic partial differential equations (PDEs). We focus on learning of the elliptic equation with a non-homogeneous Dirichlet boundary condition from independent and noisy measurements collected both inside the domain and on the boundary. Assuming that the PDE admits a strong solution in a H\"older space and using with a suitably constructed prior on the neural network weights, we prove that the posterior distribution concentrates around the exact solution at a near-minimax rate. Furthermore, the chosen prior is rate-adaptive: the posterior contracts at an (almost) optimal rate without prior knowledge of the smoothness level of the exact solution. Our results provide statistical guarantees for uncertainty quantification of PDEs via Bayesian PINNs.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
PARD-2: Target-Aligned Parallel Draft Model for Dual-Mode Speculative Decoding
arXiv:2605.08632v1 Announce Type: new
Abstract: Speculative decoding accelerates Large Language Models (LLMs) inference by using a lightweight draft model to propose candidate tokens that are verified in parallel by the target model. However, existing draft model training objectives are not directly aligned with the inference-time goal of maximizing consecutive token acceptance. To address this issue, we reformulate the draft model optimization objective, shifting the focus from token prediction accuracy to the overall acceptance length. In this paper, we build upon PARD to propose PARD-2, a dual-mode speculative decoding framework with Confidence-Adaptive Token (CAT) optimization. This approach adaptively reweights each token to better align with the verification process. Notably, PARD-2 enables a single draft model to support both target-dependent and target-independent modes. Experiments across diverse models and tasks demonstrate that PARD-2 achieves up to 6.94$\times$ lossless acceleration, surpassing EAGLE-3 by 1.9$\times$ and PARD by 1.3$\times$ on Llama3.1-8B. Our code is available at https://github.com/AMD-AGI/PARD.
Fonte: arXiv cs.CL
MLOps/Systems • Score 85
Queryable LoRA: Instruction-Regularized Routing Over Shared Low-Rank Update Atoms
arXiv:2605.08423v1 Announce Type: cross
Abstract: We present a data-adaptive method for parameter-efficient fine-tuning of large neural networks. Standard low-rank adaptation methods improve efficiency by restricting each layer update to a fixed low-rank form, but this static parameterization can be too rigid when the appropriate correction depends on the input and on the evolving depth-wise computation of the network. Our approach replaces a purely layer-local adapter with a shared queryable memory of low-rank update atoms. For each block of layers, the model forms a query from the current low-rank state and a running summary of previous blocks, uses this query to retrieve a content-dependent combination of shared update components via attention, and applies the resulting routed operator within the low-rank bottleneck. In this way, the method retains the efficiency and scalability of low-rank adaptation while allowing the effective update to vary across inputs and to share reusable structure across layers. The resulting architecture provides a principled middle ground between static LoRA-style updates and fully generated parameter updates: it remains compact and parameter-efficient while supporting dynamic, context-sensitive adaptation. Further, we incorporate instruction-regularization by augmenting routing logits with a language-induced prior over update atoms, thereby biasing the selection of low-rank transformations toward semantically relevant directions without generating unconstrained parameter updates. Experiments on noisy non-linear regression tasks and LLM fine-tuning suggest that this queryable update-memory formulation can improve final test performance and training stability compared to standard low-rank adaptation, while using a comparable number of trainable parameters.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Federated Language Models Under Bandwidth Budgets: Distillation Rates and Conformal Coverage
arXiv:2605.09986v1 Announce Type: new
Abstract: Training a language model on data scattered across bandwidth-limited nodes that cannot be centralized is a setting that arises in clinical networks, enterprise knowledge bases, and scientific consortia. We study the regime in which data must remain distributed across nodes, and ask what statistical guarantees are in principle achievable under explicit bandwidth budgets; we aim to characterize what is provably possible, not to demonstrate a deployment-ready system. Existing theory treats either training-time consistency or inference-time calibration in isolation, and none makes bandwidth a first-class statistical parameter. We analyze two protocols, Federated Probe-Logit Distillation (FPLD) for training and Federated Conformal RAG (FC-RAG) for inference, as the analytical vehicles for our results. Our first main result is an explicit high-probability KL-consistency rate for FPLD with simultaneous dependence on node count $K$, per-node sample size $n$, quantization budget $B$, probe-set size $m$, and vocabulary size $V$; bandwidth enters only through an exponentially vanishing quantization term. Our second main result is a distribution-free marginal-coverage bound for FC-RAG, whose novel retrieval-bandwidth slack $\Delta_{\mathrm{RAG}} = f_{\max}\sqrt{K^{-2}\sum_i v(B_i)}$ makes per-node retrieval bandwidth a first-class statistical parameter, with arithmetic aggregation across $K$ nodes shrinking the slack as $K^{-1/2}$ in the per-node-uniform regime. A Pinsker-type corollary composes the two bounds into an end-to-end coverage guarantee. Synthetic experiments verify the predicted scaling along the bounds' parameters; small-scale experiments on a GPT-2 testbed illustrate that the qualitative bandwidth-accuracy tradeoff survives on a real language model. A deployment-scale empirical evaluation is out of scope.
Fonte: arXiv stat.ML
Theory/Optimization • Score 85
Regret Analysis of Guided Diffusion for Black-Box Optimization over Structured Inputs
arXiv:2605.10385v1 Announce Type: new
Abstract: Guided-diffusion black-box optimization (BO) has shown strong empirical performance on structured design problems such as molecules and crystals, but its regret behavior remains poorly understood. Existing BO regret analyses typically rely on maximum information gain, non-pretrained surrogate models, or exact acquisition maximization -- assumptions that break down in modern diffusion -- BO pipelines, where pretrained diffusion models serve as powerful priors over valid structures and acquisition maximization is replaced by approximate sampling over astronomically large discrete spaces. We develop a first certificate-based expected simple-regret framework for guided-diffusion BO that avoids maximum-information-gain bounds, RKHS assumptions, and exact acquisition maximization. The central quantity in our analysis is mass lift: the increase in probability mass assigned to near-optimal designs relative to the pretrained generator. This view explains how exponential-looking finite-budget convergence and polynomial acceleration can all arise from the same mechanism. We also give practical diagnostics for estimating search exponents from finite candidate pools and a proposal-corrected resampling construction that provides a fully certified sampler instance.
Fonte: arXiv stat.ML
RL • Score 85
EvoMAS: Learning Execution-Time Workflows for Multi-Agent Systems
arXiv:2605.08769v1 Announce Type: new
Abstract: Large language model (LLM)-based multi-agent systems have shown strong potential on complex tasks through agent specialization, tool use, and collaborative reasoning. However, most automated multi-agent system design methods still follow a one-shot paradigm: a workflow is optimized or selected before execution and then reused unchanged throughout the task. This static coordination strategy is ill-suited for long-horizon tasks whose subgoals, intermediate evidence, and information needs evolve over multiple execution stages. We propose EvoMAS, a framework for execution-time multi-agent workflow construction. EvoMAS formulates workflow construction as a meta-level sequential decision problem along a single task trajectory. At each stage, it constructs an explicit task state through a Planner-Evaluator-Updater pipeline and uses a learned Workflow Adapter to instantiate a stage-specific layered workflow from a fixed pool of candidate agents. The adapter is trained with policy gradients using sparse, verifiable terminal task success as the main supervision signal, while evaluator-based process reward is analyzed separately under very-hard sparse-reward settings. Experiments on GAIA, HLE, and DeepResearcher show that EvoMAS outperforms single-agent baselines and recent automated multi-agent workflow design methods. Our analyses further show that explicit task-state construction and learned workflow adaptation provide complementary benefits. Additional results indicate that process reward is most useful when terminal success is extremely sparse, and qualitative case studies illustrate that EvoMAS adapts agent coordination as the task state evolves.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Geometry-free prediction of inertial lift forces in microfluidic devices using deep learning
arXiv:2605.08109v1 Announce Type: new
Abstract: Inertial microfluidic devices (IMDs) offer low-cost, high-throughput alternative techniques for many traditional particle- (or cell-) manipulation tasks, but simulating them requires being able to predict particle migration, and thus particle lift forces, under a variety of possible channel geometries. Recent work has demonstrated that machine learning models can be used to drastically speed up these numerical simulations, but doing so required training individual models for every unique channel cross-section type (e.g., rectangular, triangular) -- shifting the burden from the simulation step to the training step. In this paper, we develop a novel approach for predicting particle lift forces that contains no explicit geometric parameters. We train a neural network model using a new parameter set and show that while it performs comparably to existing models on channel geometries in the training set, it is able to generalize to unseen channel geometries far more effectively. We show that the lift force model developed herein can be easily transferred to particle tracing simulation software, where it is capable of predicting particle migration patterns consistent with the literature across a variety of channel designs.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Weight Pruning Amplifies Bias: A Multi-Method Study of Compressed LLMs for Edge AI
arXiv:2605.08137v1 Announce Type: new
Abstract: Weight pruning is widely advocated for deploying Large Language Models on resource-constrained IoT and edge devices, yet its impact on model fairness remains poorly understood. We conduct a controlled empirical study of three instruction-tuned models (Gemma-2-9b-it, Mistral-7B-Instruct-v0.3, Phi-3.5-mini-instruct) across three pruning methods (Random, Magnitude, Wanda) at four sparsity levels (10-70%) on 12,148 BBQ bias benchmark items with 5 random seeds, totaling 2,368,860 inference records. Our results reveal a Smart Pruning Paradox: activation-aware pruning (Wanda) preserves perplexity nearly perfectly (just 3.5% increase at 50% sparsity for Mistral-7B), yet produces the highest bias amplification, with Stereotype Reliance Score increasing 83.7% and 47-59% of previously unbiased items developing new stereotypical behaviors at 70% sparsity. Random pruning destroys language capability entirely (perplexity exceeding $10^4$ and reaching $10^8$) but produces only random-chance bias. We further show that unstructured pruning provides zero storage savings and zero inference latency reduction on real edge hardware, undermining the primary motivation for its use in IoT deployment. Of 180 dense-vs-pruned comparisons, 141 (78.3%) are significant ($p < 0.05$) with mean $|h| = 0.305$. Published quantization studies report up to 21% of responses flipping between biased and unbiased states; our pruning results show transition rates nearly three times higher (47-59%), suggesting pruning poses a categorically greater risk to alignment than quantization. These findings demonstrate that perplexity-based evaluation provides false assurance of behavioral equivalence, and that IoT deployment pipelines require bias-aware validation before deploying pruned models at the edge.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
C2L-Net: A Data-Driven Model for State-of-Charge Estimation of Lithium-Ion Batteries During Discharge
arXiv:2605.08653v1 Announce Type: new
Abstract: Accurate state-of-charge (SOC) estimation is critical for the safe and efficient operation of lithium-ion batteries in battery management systems (BMS). Although data-driven approaches can effectively capture nonlinear battery dynamics, many existing methods rely on long historical input sequences, resulting in high computational cost and introducing padding-induced positional bias at the beginning of drive cycles. To address these limitations, we propose C2L-Net, a novel context-to-latest data-driven framework for realistic online SOC estimation using only a short historical window (20 s). Unlike existing short-receptive-field or long-history models, the proposed framework explicitly separates contextual encoding from latest-measurement updating, enabling both efficient temporal modeling and rapid adaptation to dynamic battery states. The proposed model incorporates a chunk-based feature extraction mechanism that combines Theta Attention Pooling with a Fourier-based Seasonality Basis to capture local temporal patterns while reducing sequence length. A causal context encoder, integrating a gated recurrent unit (GRU) with Causal Cosine Attention, models temporal dependencies without information leakage. Furthermore, a latest-measurement decoder, inspired by recursive filtering, updates the contextual state using the most recent measurement, enhancing responsiveness to dynamic operating conditions. Extensive experiments on a public lithium-ion battery drive-cycle dataset under multiple fixed-temperature conditions demonstrate that the proposed method achieves state-of-the-art or competitive accuracy while significantly improving computational efficiency. In particular, C2L-Net achieves up to 60 times faster inference and requires fewer parameters than recent data-driven baselines, while maintaining robust performance across unseen driving profiles.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Human-Inspired Memory Architecture for LLM Agents
arXiv:2605.08538v1 Announce Type: new
Abstract: Current LLM agents lack principled mechanisms for managing persistent memory across long interaction horizons. We present a biologically-grounded memory architecture comprising six cognitive mechanisms: (1) sleep-phase consolidation, (2) interference-based forgetting, (3) engram maturation, (4) reconsolidation upon retrieval, (5) entity knowledge graphs, and (6) hybrid multi-cue retrieval. Each mechanism addresses a specific failure mode of naive memory accumulation. We introduce a synthetic calibration methodology that derives all pipeline thresholds without benchmark data exposure, eliminating a common source of evaluation leakage. We evaluate on two benchmarks. First, a VSCode issue-tracking dataset (13K issues, 120K events) where deduplication-based consolidation achieves 97.2% retention precision with 58% store reduction (+21.8 pp over baseline). Second, the LongMemEval personal-chat benchmark where we conduct the first streaming M-tier evaluation (475 sessions, ~540K unique turns). At a 200K-token context budget, our pipeline matches raw retrieval accuracy (70.1% vs. 71.2%, overlapping 95% CI) while exposing a tunable accuracy/store-size operating curve. At S-tier scale (50 sessions), dedup-based consolidation yields a +13.3 pp improvement in preference recall.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Block-Wise Differentiable Sinkhorn Attention: Tail-Refinement Gradients with a Gap-Aware Dustbin Bridge
arXiv:2605.08123v1 Announce Type: new
Abstract: We study long-context balanced entropic optimal transport (OT) attention on TPU hardware through a stopped-base, fixed-depth tail-refinement surrogate. After a stopped $T$-step Sinkhorn solve, we unroll a short refinement tail and differentiate that surrogate exactly. For the production $R=2$ case, the backward pass contains four staircase plan factors. We prove an exact one-reference-tile schedule: the $R=2$ score cotangent is a single reference plan tile times an explicit modifier field built from vector cotangents and dual differences. This yields block-wise cost $O((T+R)LW)$, $O(Ld)$ input storage, and $O(L)$ additional HBM usage for fixed head dimension $d$ and band width $W$. We also formalize the current \texttt{dustbin\_block} path as the same balanced surrogate on an augmented support, so the schedule lifts to the gap-aware transport path used in our TPU runs. We provide a local surrogate-bias bound, an a posteriori bias certificate, and a projective contraction certificate for strictly positive active blocks. On synthetic masked problems, the optimized kernel matches exact autodiff of the same centered surrogate to within $10^{-5}$--$10^{-10}$. On TPU v6e-8, a four-configuration Pfam screen completes end-to-end, and a promoted balanced $R=2$ run sustains roughly $8.5$ examples per second through a three-hour budget, reaching step $1437$. Held-out Pfam test shards improve reconstruction from $3.17$ to $0.99$ and sparse CE from $5.86$ to $5.69$ relative to step $0$. These results support exact fixed-depth backward theory, a theorem-matching gap-aware bridge, and trainability evidence for the production path.
Fonte: arXiv cs.LG
Privacy/Security/Fairness • Score 85
FairHealth: An Open-Source Python Library for Trustworthy Healthcare AI in Low-Resource Settings
arXiv:2605.08198v1 Announce Type: new
Abstract: We present FairHealth, an open-source Python library that provides a unified, modular framework for trustworthy machine learning in healthcare applications, with particular focus on low-resource and low-income country (LMIC) settings such as Bangladesh. FairHealth addresses four critical gaps in existing healthcare AI toolkits: (1) the absence of integrated fairness auditing for biosignals and clinical tabular data; (2) the lack of privacy-preserving federated learning tools compatible with standard ML workflows; (3) missing explainability tools tailored for low-bandwidth clinical decision support; and (4) no existing toolkit covering Global South healthcare datasets. Built from five peer-reviewed research contributions, FairHealth provides six modules covering federated learning with homomorphic encryption (fairhealth.federated), intersectional fairness metrics (fairhealth.fairness), hybrid fuzzy-SHAP explainability (fairhealth.explain), multilingual dengue triage (fairhealth.lowresource), equitable disaster aid allocation (fairhealth.equity), and public dataset loaders (fairhealth.datasets). All datasets used are publicly available without institutional data use agreements. FairHealth is installable via pip install fairhealth(PyPI: pypi.org/project/fairhealth/) and available at https://github.com/Farjana-Yesmin/fairhealth.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
cuRegOT: A GPU-Accelerated Solver for Entropic-Regularized Optimal Transport
arXiv:2605.08793v1 Announce Type: cross
Abstract: Optimal transport (OT) has emerged as a fundamental tool in modern machine learning, yet its computational cost remains a significant bottleneck for large-scale applications. While harnessing the massive parallelism of modern GPU hardware is critical for efficiency, the de facto standard Sinkhorn algorithm, despite its ease of parallelization, often suffers from slow convergence in challenging problems. More recently, the sparse-plus-low-rank quasi-Newton method offers a balance between convergence rate and per-iteration complexity; however, its efficiency on GPUs is severely hindered by the serial nature of sparse matrix symbolic analysis and irregular memory access patterns. To bridge this gap, we present cuRegOT, a high-performance GPU solver tailored for entropic-regularized OT. We introduce a suite of algorithmic and architectural optimizations, including an amortized symbolic analysis strategy to mitigate CPU bottlenecks, an asynchronous Sinkhorn iterates generation mechanism, and a fused kernel for bandwidth-efficient gradient evaluation. These strategies are backed by rigorous theoretical guarantees ensuring algorithmic convergence. Extensive numerical experiments demonstrate that cuRegOT achieves significant speedups over state-of-the-art GPU-based solvers across a variety of benchmark tasks.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
EdgeFlowerTune: Evaluating Federated LLM Fine-Tuning Under Realistic Edge System Constraints
arXiv:2605.08636v1 Announce Type: new
Abstract: Federated fine-tuning offers a promising paradigm for adapting large language models (LLMs) on edge devices by leveraging the rich, diverse, and continuously generated data from smartphones and IoT devices without compromising user data privacy. Such edge-side adaptation can improve model personalization, robustness, and responsiveness to local contexts. However, the practical feasibility of federated LLM fine-tuning on real edge devices remains unclear, as most existing work focuses on cross-silo or simulation-based settings, overlooking the resource and runtime constraints that determine whether a method is deployable on real edge systems. We present EdgeFlowerTune, a deployment-oriented benchmark for federated LLM fine-tuning under realistic edge-system constraints. EdgeFlowerTune jointly evaluates model quality and system costs, including communication, wall-clock latency, memory usage, energy consumption, and robustness to dynamic edge conditions. To compare methods in terms of effectiveness, efficiency, and robustness, EdgeFlowerTune introduces three complementary protocols: Quality-under-Budget, Cost-to-Target, and Robustness. We instantiate EdgeFlowerTune as a real-device platform built on Flower and MobileFineTuner, spanning commercial Android smartphones and NVIDIA edge development boards. Our benchmark results show that accuracy-only evaluation can lead to misleading conclusions: methods with similar final quality may differ substantially in deployability once realistic system constraints are considered. EdgeFlowerTune provides a reproducible benchmark for system-aware evaluation of federated LLM fine-tuning at the edge.
Fonte: arXiv cs.CL
MLOps/Systems • Score 85
Text-Guided Multi-Scale Frequency Representation Adaptation
arXiv:2605.08181v1 Announce Type: new
Abstract: Parameter-efficient fine-tuning methods introduce a small number of training parameters, enabling pre-trained models to adapt rapidly to new data distributions. While these methods have shown promising results, they exhibit notable limitations. First, most existing methods operate in the signal space domain, which results in substantial information redundancy. Second, most existing methods utilize fixed prompts or adaptation layers, failing to fully account for the multi-scale characteristics of signals. To address these challenges, we propose the Multi-Scale Frequency Adapter (FreqAdapter), which integrates textual information and performs multi-scale fine-tuning of signals in the frequency domain. Additionally, we introduce a multi-scale adaptation strategy to optimize receptive fields across different frequency ranges, further enhancing the model's representational capacity. Extensive experiments on multimodal models, including CLIP and LLaVA, demonstrate that FreqAdapter significantly improves both performance and efficiency. FreqAdapter improves performance with minimal cost and fast convergence within one epoch. Code is available at https://github.com/Kelvin-ywc/FreqAdapter.
Fonte: arXiv cs.CV
Theory/Optimization • Score 85
Rennala MVR: Improved Time Complexity for Parallel Stochastic Optimization via Momentum-Based Variance Reduction
arXiv:2605.08871v1 Announce Type: cross
Abstract: Large-scale machine learning models are trained on clusters of machines that exhibit heterogeneous performance due to hardware variability, network delays, and system-level instabilities. In such environments, time complexity rather than iteration complexity becomes the relevant performance metric for optimization algorithms. Recent work by Tyurin and Richt\'{a}rik (2023) established the first time complexity analysis for parallel first-order stochastic optimization, proposing Rennala SGD as a time-optimal method for smooth nonconvex optimization. However, Rennala SGD is fundamentally a modification of SGD, and variance reduction techniques are known to improve the iteration complexity of SGD. In this work, we investigate whether variance reduction can also improve time complexity in heterogeneous systems. We show that, under a mean-squared smoothness assumption, variance reduction can improve time complexity in relevant parameter regimes. To this end, we propose Rennala MVR, a variance-reduced extension of Rennala SGD based on momentum-based variance reduction, and analyze its oracle and time complexity. We establish lower bounds for time complexity under these assumptions. On a stochastic quadratic benchmark, experiments with the exact method support the theory, while neural-network experiments with a practical inexact variant show similar empirical gains over Rennala SGD.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
MIND-Skill: Quality-Guaranteed Skill Generation via Multi-Agent Induction and Deduction
arXiv:2605.08670v1 Announce Type: new
Abstract: Large language model (LLM) powered AI agents have emerged as a promising paradigm for autonomous problem-solving, yet they continue to struggle with complex, multi-step real-world tasks that demand domain-specific procedural knowledge. Reusable agent skills, which encapsulate successful problem-solving strategies, offer a natural remedy by enabling agents to build on prior experience. However, curating such skills has largely remained a manual endeavor, requiring human experts to distill rich domain knowledge into actionable guidelines. In this work, we present $\textbf{M}$ulti-agent $\textbf{IN}$duction and $\textbf{D}$eduction for $\textbf{Skill}$s ($\textbf{MIND-Skill}$), a framework that automatically induces generalizable skills from successful trajectories with robust quality guarantees. MIND-Skill consists of an induction agent which is tasked to abstract reusable skills from successful trajectories, and a deduction agent which aims to reconstruct trajectories by following the induced skills. To guarantee the quality of the generated skills, we introduce a reconstruction loss that compares input and reconstructed trajectories, an outcome loss that enforces the correctness of the reconstructed trajectories, and a rubric loss that assesses the documentation quality and regularizes the abstraction level of the generated skills according to predefined criteria. These textual losses are jointly optimized with TextGrad, and the resulting skills are evaluated on held-out tasks unseen during optimization. Experiments on AppWorld and BFCL-v3 show that MIND-Skill consistently outperforms concurrent skill generation methods.
Fonte: arXiv cs.AI
RL • Score 85
MemQ: Integrando Q-Learning em Agentes de Memória Autoevolutivos sobre DAGs de Proveniência
arXiv:2605.08374v1 Tipo de Anúncio: novo
Resumo: A memória episódica permite que agentes LLM acumulem e recuperem experiências, mas os métodos atuais tratam cada memória de forma independente, ou seja, avaliando a qualidade da recuperação de forma isolada, sem considerar as cadeias de dependência através das quais as memórias possibilitam a criação de memórias futuras. Introduzimos o MemQ, que aplica traços de elegibilidade TD($\lambda$) aos valores Q de memória, propagando crédito para trás através de um DAG de proveniência que registra quais memórias foram recuperadas quando cada nova memória foi criada. O peso do crédito decai à medida que $(\gamma\lambda)^d$ com a profundidade do DAG $d$, substituindo a distância temporal por proximidade estrutural. Formalizamos o cenário como um MDP de Contexto Exógeno, cuja transição fatorada desacopla o fluxo de tarefas exógeno do armazenamento de memória endógeno. Em seis benchmarks, abrangendo interação com o OS, chamada de funções, geração de código, raciocínio multimodal, raciocínio incorporado e QA em nível de especialista, o MemQ alcança a maior taxa de sucesso em todos os seis na avaliação de generalização e aprendizado em tempo de execução, com ganhos maiores em tarefas de múltiplos passos que produzem cadeias de proveniência profundas e relevantes (até +5.7~pp) e menores em classificação de um único passo (+0.77~pp), onde atualizações de um único passo já são suficientes. Estudamos ainda como $\gamma$ e $\lambda$ interagem com a estrutura do EC-MDP, fornecendo orientações fundamentadas para seleção de parâmetros e pesquisas futuras. O código estará disponível em breve.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
DARE: Diffusion Language Model Activation Reuse for Efficient Inference
arXiv:2605.08134v1 Announce Type: new
Abstract: Diffusion Large Language Models (dLLMs) have emerged as a promising alternative to auto-regressive (AR) models, offering greater expressive capacity and potential for parallel generation and faster inference. However, open-source dLLMs remain immature, lagging behind AR models in both efficiency and quality.
We identify an underexplored property of dLLMs: *token-wise redundancy* in bi-directional self-attention. Self-attention activations are highly correlated across tokens, and temporal changes in query representations can predict redundancy in corresponding key, value, and output activations.
We introduce DARE, with two complementary mechanisms: DARE-KV, which reuses cached key-value (KV) activations, and DARE-O, which reuses output activations to reduce redundant computation while preserving quality.
DARE achieves up to 1.20x per-layer latency reduction and reuses up to 87% of attention activations, with negligible degradation on reasoning and code-generation benchmarks. DARE-KV and DARE-O incur average performance drops of only 2.0% and 1.2%, respectively. Combined with techniques such as prefix caching and Fast-dLLM, DARE provides additive gains without retraining.
These results establish token-wise reuse as an effective strategy for improving the efficiency of diffusion-based LLMs while preserving generation fidelity. Code: https://github.com/enyac-group/DARE
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Communication Dynamics Neural Networks: FFT-Diagonalized Layers for Improved Hessian Conditioning at Reduced Parameter Count
arXiv:2605.08171v1 Announce Type: new
Abstract: Background and motivation. The Communication Dynamics (CD) framework, introduced in two earlier papers for atomic-energy prediction and field-induced superconductivity, treats each physical channel as a (2l+1)-vertex polygon whose discrete Fourier transform yields its energy spectrum. This paper applies the same circulant-spectral machinery to neural-network design.
Layer construction. CDLinear is a block-circulant linear layer with block size B = 2l+1 and 1/B the parameter count of a dense layer of equal input/output dimensions. Three properties follow from the construction. (i) The Hessian of mean-squared loss with respect to the weights is diagonalized by the discrete Fourier transform, with eigenvalues |F[Xj](k)|^2 read directly from the input statistics (Theorem 1). (ii) Under input pre-whitening, the population Hessian condition number satisfies kappa = 1 exactly, with the empirical condition number bounded by 1+O(sqrt(B/N)) on N samples (Theorem 2). (iii) The Shannon noise rate alpha_CD = 0.0118 calibrated in the parent CD papers from the Na D-doublet specifies a transferable, non-arbitrary dropout rate.
Empirical evaluation. A CDLinear MLP at B = 4 achieves 97.50% +/- 0.23% test accuracy with 2,380 parameters versus 98.15% +/- 0.47% for a parameter-matched dense MLP at 8,970 parameters, a 3.8x parameter reduction at 0.65% accuracy cost, within one standard deviation of the seed-to-seed spread. The CD-MLP mean Hessian condition number kappa = 1.9x10^4 is 310x smaller than the dense baseline kappa = 5.9x10^6, in quantitative agreement with Theorem 2.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
CERSA: Cumulative Energy-Retaining Subspace Adaptation for Memory-Efficient Fine-Tuning
arXiv:2605.08174v1 Announce Type: new
Abstract: To mitigate the memory constraints associated with fine-tuning large pre-trained models, existing parameter-efficient fine-tuning (PEFT) methods, such as LoRA, rely on low-rank updates. However, such updates fail to fully capture the rank characteristics of the weight modifications observed in full-parameter fine-tuning, resulting in a performance gap. Furthermore, LoRA and other existing PEFT methods still require substantial memory to store the full set of frozen weights, limiting their efficiency in resource-constrained settings. To addres these limitations, we introduce Cumulative Energy-Retaining Subspace Adaptation (CERSA), a novel fine-tuning paradigm that leverages singular value decomposition (SVD) to retain only the principal components responsible for 90% to 95% of the spectral energy. By fine-tuning low-rank representations derived from this principal subspace, CERSA significantly reduces memory consumption. We conduct extensive evaluations of CERSA across models of varying scales and domains, including image recognition, text-to-image generation, and natural language understanding. Empirical results demonstrate that CERSA consistently outperforms state-of-the-art PEFT methods while achieving substantially lower memory requirements. The code will be publicly released.
Fonte: arXiv cs.LG
NLP/LLMs • Score 90
ReST-KV: Robust KV Cache Eviction with Layer-wise Output Reconstruction and Spatial-Temporal Smoothing
arXiv:2605.08840v1 Announce Type: new
Abstract: Large language models (LLMs) face growing challenges in efficient generative inference due to the increasing memory demands of Key-Value (KV) caches, especially for long sequences. Existing eviction methods typically retain KV pairs with high attention weights but overlook the impact of attention redistribution caused by token removal, as well as the spatial-temporal dynamics in KV selection. In this paper, we propose ReST-KV, a robust KV eviction method that combines layer-wise output Reconstruction and Spatial-Temporal smoothing to provide a more comprehensive perspective for the KV cache eviction task. Specifically, ReST-KV formulates KV cache eviction as an optimization problem that minimizes output discrepancies through efficient layer-wise reconstruction. By directly modeling how each token's removal affects the model output, our method naturally captures attention redistribution effects, going beyond simplistic reliance on raw attention weights. To further enhance robustness, we design exponential moving average smoothing to handle temporal variations and an adaptive window-based mechanism to capture spatial patterns. Our method, ReST-KV, significantly advances performance on long-context benchmarks. It surpasses state-of-the-art baselines by 2.58% on LongBench and 15.2% on RULER. Additionally, ReST-KV consistently outperforms existing methods on Needle-in-a-Haystack and InfiniteBench, all while achieving a remarkable 10.61$\times$ reduction in decoding latency at 128k context length. The code is publicly available at https://github.com/an-yongqi/rest-kv to facilitate reproducibility and further research.
Fonte: arXiv cs.CL
MLOps/Systems • Score 85
Core-Halo Decomposition: Decentralizing Large-Scale Fixed-Point Problems
arXiv:2605.08681v1 Announce Type: new
Abstract: We study solving large-scale fixed-point equation \(x^\star=\bar F(x^\star)\) with decomposition. Standard strict decomposition assigns each agent a disjoint block and evaluates updates using only owned coordinates. For most operators, however, a block update may depend on variables outside the block. Truncating these dependencies by strict decomposition changes the mean operator and creates structural bias that cannot be removed by more samples, smaller stepsizes, or additional consensus. We therefore propose Core-Halo decomposition, which separates write ownership from read-only evaluation context: each agent updates its own core and reads from an overlapping halo. By aligning the Core-Halo decomposition with the block-dependence structure of $\bar F$, the original fixed-point problem can be implemented faithfully in a decentralized multi-agent system. We further characterize the fundamental obstruction faced by strict decomposition through a Bellman closure condition and a blockwise bias lower bound, showing that local-only updates can alter the original fixed-point operator. Finally, we conduct extensive experiments across a range of application settings, and demonstrate that Core-Halo achieves near-centralized performance while retaining the parallelism benefits of decentralization.
Fonte: arXiv stat.ML
Privacy/Security/Fairness • Score 85
A Simulated Federated Analysis of MS-Induced Brain Lesions
arXiv:2605.08223v1 Announce Type: new
Abstract: Federated techniques such as federated learning and federated analysis have emerged as a powerful paradigm for enabling multi-center research on sensitive clinical data while preserving patient privacy. In this study, we introduce a simulation framework that emulates a real-world federated research project focused on the analysis of multiple sclerosis (MS) patient data. The project comprises two components: an image segmentation task and a clinical data analysis task, where federated variants of survival analysis and Principal Component Analysis (PCA) are employed. To capture the complexity and heterogeneity of real clinical datasets, we construct a federation of high-fidelity synthetic cohorts designed to mirror MS-related clinical and demographic characteristics, while the imaging component leverages publicly available real-world datasets. Our simulation replicates key elements of authentic federated workflows, including distributed data governance, site-specific preprocessing, model training across isolated nodes, and the secure aggregation of analytical outputs. This framework provides a realistic testbed for developing, evaluating, and benchmarking federated learning methods in the context of MS research.
Fonte: arXiv cs.LG
Multimodal • Score 85
From Historical Tabular Image to Knowledge Graphs: A Provenance-Aware Modular Pipeline
arXiv:2605.08222v1 Announce Type: new
Abstract: Handwritten archival tables contain rich historical information, yet transforming them into structured representations, such as Knowledge Graphs, requires integrating table structure recognition, handwriting recognition, and semantic interpretation - a complex multimodal process. End-to-end AI implementations can obscure these steps, resulting in opaque algorithmic operations that hinder human oversight, critical assessment, and trust. To address this, we present a modular, provenance-aware pipeline to convert handwritten tabular images into KGs supporting human-AI collaboration. The pipeline decomposes the workflow into three stages - table reconstruction, information extraction, and KG construction - while exposing intermediate representations for inspection, evaluation, and correction. A key contribution of our approach is the systematic integration of data provenance at every stage, ensuring that all extracted entities and literals remain traceable to their visual and textual origins. The proposed pipeline is demonstrated through a number of experiments on real-world archival material concerning military careers. The results across three different table reconstruction variants highlight the importance of modularisation. By coupling modularity with data provenance, our work advances transparent and collaboratively controllable image-to-KG pipelines for complex historical data.
Fonte: arXiv cs.CV
Theory/Optimization • Score 85
Learning stochastic multiscale models through normalizing flows
arXiv:2605.09718v1 Announce Type: new
Abstract: Many systems in physics, engineering, and biology exhibit multiscale stochastic dynamics, where low-dimensional slow variables evolve under the influence of high-dimensional fast processes. In practice, observations are often limited to a single trajectory of the slow component, while the fast dynamics remain unobserved, making statistical learning challenging. Approaches based on partial differential equations (PDE), such as Fokker-Planck formulations, aim to characterize the evolution of probability densities, typically requiring dense space-time data or grid-based solvers. In contrast, we adopt a trajectory-based perspective and develop a data-driven framework for learning effective stochastic dynamics from a single observed path. We model the dynamics by coupled multiscale stochastic differential equations (SDEs) and first obtain a principled model reduction through stochastic averaging. Unlike generic model reduction techniques such as PCA, this respects the dynamical structure of the original system and explicitly incorporates the interaction between slow and fast scales. A central challenge, however, is that the reduced model depends on the invariant distribution of the fast process, which is a solution to an intractable and often unknown PDE. We introduce a novel learning framework that parameterizes the invariant distribution using normalizing flows, enabling expressive density modeling in the latent fast-variable space. The flow is trained end-to-end by optimizing a penalized likelihood objective induced by the reduced stochastic dynamics. Furthermore, we develop a Bayesian variational inference procedure for uncertainty quantification, employing a second normalizing flow to approximate the posterior distribution over model parameters. This yields a scalable approach to capturing epistemic uncertainty in multiscale systems.
Fonte: arXiv stat.ML
RL • Score 85
OracleTSC: Regularização de Hurdle de Recompensa e Incerteza Informada pela Oracle para Controle de Sinais de Trânsito
arXiv:2605.08516v1 Tipo de Anúncio: novo
Resumo: A tomada de decisão transparente é essencial para que os sistemas de controle de sinais de trânsito (TSC) ganhem a confiança do público. No entanto, os métodos tradicionais de TSC baseados em aprendizado por reforço funcionam como caixas-pretas com interpretabilidade limitada. Embora grandes modelos de linguagem (LLMs) possam fornecer raciocínio em linguagem natural, o ajuste fino por reforço para TSC continua instável porque o feedback é escasso e atrasado, enquanto a maioria das ações produz apenas mudanças marginais nas métricas de congestionamento. Introduzimos o OracleTSC, que estabiliza o TSC baseado em LLM por meio de dois mecanismos: (1) um mecanismo de hurdle de recompensa que filtra sinais de aprendizado fracos subtraindo um limite calibrado das recompensas ambientais, e (2) regularização de incerteza que maximiza a probabilidade da resposta selecionada para incentivar decisões consistentes entre as saídas amostradas. Experimentos no benchmark LibSignal mostram que o OracleTSC permite que um modelo compacto LLaMA3-8B melhore substancialmente a eficiência do tráfego, alcançando uma redução de 75% no tempo de viagem e uma diminuição de 67% no comprimento da fila em comparação com a linha de base pré-treinada, mantendo a interpretabilidade por meio de explicações em linguagem natural. O OracleTSC também demonstra forte generalização entre interseções: uma política treinada em uma interseção transfere para uma interseção estruturalmente diferente com 17% menos tempo de viagem e 39% menos comprimento de fila sem ajuste fino adicional. Esses resultados sugerem que a modelagem de recompensa ciente da incerteza pode melhorar a estabilidade e a eficácia do ajuste fino por reforço para TSC.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Auto-Rubric como Recompensa: De Preferências Implícitas a Critérios Gerativos Multimodais Explícitos
arXiv:2605.08354v1 Tipo de Anúncio: novo
Resumo: Alinhar modelos generativos multimodais com preferências humanas exige sinais de recompensa que respeitem a estrutura composicional e multidimensional do julgamento humano. As abordagens prevalentes de RLHF reduzem essa estrutura a rótulos escalares ou pareados, colapsando preferências sutis em proxies paramétricos opacos e expondo vulnerabilidades à manipulação de recompensas. Embora métodos recentes de Rubrics-as-Reward (RaR) tentem recuperar essa estrutura por meio de critérios explícitos, gerar rubricas que sejam simultaneamente confiáveis, escaláveis e eficientes em termos de dados continua sendo um problema em aberto. Introduzimos Auto-Rubric como Recompensa (ARR), uma estrutura que reformula a modelagem de recompensas de otimização de peso implícito para decomposição explícita baseada em critérios. Antes de qualquer comparação pareada, o ARR externaliza o conhecimento de preferência internalizado de um VLM como rubricas específicas de prompt, traduzindo a intenção holística em dimensões de qualidade independentes e verificáveis. Essa conversão da estrutura de preferência implícita em restrições inspecionáveis e interpretáveis suprime substancialmente os viéses de avaliação, incluindo viés posicional, permitindo tanto a implantação zero-shot quanto o condicionamento few-shot com supervisão mínima. Para estender esses ganhos ao treinamento generativo, propomos a Otimização de Política de Rubrica (RPO), que destila a avaliação estruturada multidimensional do ARR em uma recompensa binária robusta, substituindo a regressão escalar opaca por decisões de preferência condicionadas por rubrica que estabilizam os gradientes de política. Em benchmarks de geração de texto para imagem e edição de imagem, o ARR-RPO supera modelos de recompensa pareados e juízes VLM, demonstrando que externalizar explicitamente o conhecimento de preferência implícita em rubricas estruturadas alcança um alinhamento multimodal mais confiável e eficiente em termos de dados, revelando que o gargalo é a ausência de uma interface fatorada, não um déficit de conhecimento.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Generalized Category Discovery in Federated Graph Learning
arXiv:2605.08178v1 Announce Type: new
Abstract: Federated Graph Learning (FGL) enables collaborative learning over distributed graph data, yet existing approaches largely rely on a closed-world assumption, limiting their applicability in dynamic environments where novel categories continuously emerge. To bridge this gap, we target the practical scenario of Federated Graph Generalized Category Discovery (FGGCD), aiming to collaboratively discover novel categories across decentralized graph clients while retaining knowledge of known categories. We observe that FGGCD introduces two fundamental challenges: (1) the Neighborhood Absorption Effect, where structural fragmentation leads to biased neighborhood aggregation, causing novel nodes to be misclassified as known categories; and (2) Global Semantic Inconsistency, where the aforementioned local biases propagate to the server and are amplified by heterogeneous subgraph distributions, hindering cross-client knowledge integration. To address these issues, we propose GCD-FGL, an FGL framework for GCD that integrates a client-side Topology-Reliable Semantic Alignment and Discovery process to mitigate the neighborhood absorption effect, and a server-side Hierarchical Prototype Alignment strategy to resolve global semantic inconsistency. Extensive experiments on five real-world graph datasets demonstrate that GCD-FGL consistently outperforms state-of-the-art baselines, achieving an average absolute gain of +4.86 in HRScore.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Learngene Search Across Multiple Datasets for Building Variable-Sized Models
arXiv:2605.08209v1 Announce Type: new
Abstract: Deep learning methods are widely used under diverse resource constraints, resulting in models of varying sizes, such as the Vision Transformer (ViT) series. Deploying these models typically requires costly pretraining and finetuning. The Learngene paradigm addresses this issue by extracting transferable components, called learngenes, from a pretrained ancestry model (Ans-Net) to initialize variable-sized descendant models (Des-Nets).Existing learngene extraction methods rely on a single dataset, limiting downstream performance. To address this limitation, we propose Learngene Search Across Multiple Datasets for Building Variable-Sized Models (LSAMD). LSAMD expands the Ans-Net into a searchable super Ans-Net with dataset-specific blocks and dataset adapters (DADs). During training, LSAMD searches for an optimal architecture path for each dataset. The base blocks most frequently selected across datasets are extracted as learngenes for initializing Des-Nets.Experiments on multiple datasets show that LSAMD achieves performance comparable to pretrain-finetune methods while significantly reducing storage and training costs.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
ExecuTorch -- A Unified PyTorch Solution to Run AI Models On-Device
arXiv:2605.08195v1 Announce Type: new
Abstract: Local execution of AI on edge devices is important for low latency and offline operation. However, deploying models on diverse hardware remains fragmented, often requiring model conversion or complete reimplementation outside the PyTorch ecosystem where the model was originally authored. We introduce ExecuTorch, a unified PyTorch-native deployment framework for edge AI. ExecuTorch enables seamless deployment of machine learning models across heterogeneous compute environments. It scales from embedded microcontrollers to complex system-on-chips (SoCs) with dedicated accelerators, powering devices ranging from wearables and smartphones to large compute clusters. ExecuTorch preserves PyTorch semantics while allowing customization, support for optimizations like quantization, and pluggable execution "backends". These features together enable fast experimentation, allowing researchers to validate deployment behavior entirely within PyTorch, bridging the gap between research and production.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Structured Recurrent Mixers for Massively Parallelized Sequence Generation
arXiv:2605.08696v1 Announce Type: new
Abstract: Over the last two decades, language modeling has experienced a shift from predominantly recurrent architectures that process tokens sequentially during training and inference to non-recurrent models that process sequence elements in parallel during training, which results in greater training efficiency and stability at the expense of lower inference throughput. Here we introduce the Structured Recurrent Mixer, an architecture that allows for algebraic conversion between a sequence parallel representation at train time and a recurrent representation at inference, notably without the need for specialized kernels or device-specific memory management. We show experimentally that this dual representation allows for greater training efficiency, higher input information capacity, and larger inference throughput and concurrency when compared to other linear complexity models. We postulate that recurrent models are poorly suited to extended sequence length scaling for information-rich inputs typical of language, but are well suited to scaling in the sample (batch) dimension due to their constant memory per sample. We provide Mojo/MAX inference implementations of SRMs exhibiting 12x the throughput and 170x the concurrency of similarly powerful Transformers inferenced on vLLM, increases characteristic of Pytorch implementations resulting in a 30\% increase in compute-constant GSM8k Pass@k. We conclude by demonstrating that SRMs are effective reinforcement learning training candidates.
Fonte: arXiv cs.CL
RL • Score 85
Learning Visual Feature-Based World Models via Residual Latent Action
arXiv:2605.07079v1 Announce Type: new
Abstract: World models predict future transitions from observations and actions. Existing works predominantly focus on image generation only. Visual feature-based world models, on the other hand, predict future visual features instead of raw video pixels, offering a promising alternative that is more efficient and less prone to hallucination. However, current feature-based approaches rely on direct regression, which leads to blurry or collapsed predictions in complex interactions, while generative modeling in high-dimensional feature spaces still remains challenging. In this work, we discover that a new type of latent action representation, which we refer to as *Residual Latent Action* (RLA), can be easily learned from DINO residuals. We also show that RLA is predictive, generalizable, and encodes temporal progression. Building on RLA, we propose *RLA World Model* (RLA-WM), which predicts RLA values via flow matching. RLA-WM outperforms both state-of-the-art feature-based and video-diffusion world models on simulation and real-world datasets, while being orders of magnitude faster than video diffusion. Furthermore, we develop two robot learning techniques that use RLA-WM to improve policy learning. The first one is a minimalist world action model with RLA that learns from actionless demonstration videos. The second one is the first visual RL framework trained entirely inside a world model learned from offline videos only, using a video-aligned reward and no online interactions or handcrafted rewards. Project page: https://mlzxy.github.io/rla-wm
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
MEMOREPAIR: Barrier-First Cascade Repair in Agentic Memory
arXiv:2605.07242v1 Announce Type: new
Abstract: Agentic memory evolves across tasks into durable derived artifacts: summaries, cached outputs, embeddings, learned skills, and executable tool procedures. When a source artifact is deleted, corrected, or invalidated by tool or API migration, descendants derived from that source can remain visible and steer future actions with stale support. We formalize this failure mode as the cascade update problem, where repair targets the visible derived state of the memory store. We present MemoRepair, a barrier-first cascade-repair contract for agentic memory. A repair event induces a controlled transition from invalidated descendant state to validated successor state: affected descendants are withdrawn before repair, successors are constructed from retained support and staged repaired predecessors under the current interface, and republication is restricted to validated predecessor-closed successors. This contract induces a scalarized repair-selection problem for a fixed repair-cost tradeoff. We show that the induced publication problem reduces to maximum-weight predecessor closure and can be solved exactly by a single s-t min-cut. Experiments on ToolBench and MemoryArena show that, with complete influence provenance, MemoRepair reduces invalidated-memory exposure from 69.8-94.3% under systems without cascade repair to 0%. Compared with exhaustive Repair all, it recovers 91.1-94.3% of validated successors while reducing normalized repair-operator cost from 1.00 to 0.57-0.76.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Conformal Agent Error Attribution
arXiv:2605.06788v1 Announce Type: new
Abstract: When multi-agent systems (MAS) fail, identifying where the decisive error occurred is the first step for automated recovery to an earlier state. Error attribution remains a fundamental challenge due to the long interaction traces that large language model-based MAS generate. This paper presents a framework for error attribution based on conformal prediction (CP) which provides finite-sample, distribution-free coverage guarantees. We introduce new algorithms for filtration-based CP designed for sequential data such as agent trajectories. Unlike existing CP algorithms, our approach predicts sets that are contiguous sequences to enable efficient recovery and debugging. We verify our theoretical guarantees on a variety of agents and datasets, show that errors can be precisely isolated, then use prediction sets to rollback MAS to correct their own errors. Our overall approach is model-agnostic, and offers a principled uncertainty layer for MAS error attribution. We release code at https://github.com/layer6ai-labs/conformal-agent-error-attribution.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Beyond Factor Aggregation: Gauge-Aware Low-Rank Server Representations for Federated LoRA
arXiv:2605.06733v1 Announce Type: new
Abstract: Federated LoRA enables parameter-efficient adaptation of large language models under decentralized data and limited client resources.However, directly averaging LoRA factors is representation-dependent: the same intrinsic update admits infinitely many gauge-equivalent factorizations, so factor-level aggregation can change under arbitrary coordinate choices while the underlying update remains unchanged. This reveals a semantic mismatch in existing federated LoRA aggregation rules. We propose \textbf{GLoRA}, a gauge-aware server representation for federated LoRA.Instead of aggregating raw factors, GLoRA estimates a consensus update subspace from client projectors and aggregates client updates in shared reference coordinates, thereby representing semantic update aggregation entirely in low-rank form. To support heterogeneous client capacities, GLoRA further provides a rank-compatible readout that instantiates adapters of different ranks from the same server state without dense update reconstruction. Experiments on GLUE and SuperNI show that GLoRA consistently outperforms federated LoRA baselines under data, resource, and task heterogeneity, including heterogeneous client ranks, sparse participation, larger backbones, and unseen-task evaluation. GLoRA also achieves a favorable efficiency--performance trade-off, suggesting that effective federated LoRA requires not merely averaging low-rank factors, but defining a semantically meaningful server-side representation for aggregation.
Fonte: arXiv cs.LG
NLP/LLMs • Score 90
Learning and Reusing Policy Decompositions for Hierarchical Generalized Planning with LLM Agents
arXiv:2605.06957v1 Announce Type: new
Abstract: We present a dynamic policy-learning approach that combines generalized planning and hierarchical task decomposition for LLM-based agents. Our method, Hierarchical Component Learning for Generalized Policies (HCL-GP ), learns parameterized policies that generalize across task instances and automatically extracts reusable components from successful executions, organizing them into a component library for compositional policy generation. We address three challenges: (1) learning components through automated decomposition, (2) generalizing components to maximize reuse, and (3) efficient retrieval via semantic search. Evaluated on the AppWorld benchmark, our approach achieves 98.2% accuracy on normal tasks and 97.8% on challenge tasks with unseen applications, improving 15.8 points over static synthesis on challenging scenarios. For open-source models, dynamic reuse enables 62.5% success versus near-zero without reuse. This demonstrates that classical planning concepts can be effectively integrated with LLM agents for improved accuracy and efficiency.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Weblica: Scalable and Reproducible Training Environments for Visual Web Agents
arXiv:2605.06761v1 Announce Type: new
Abstract: The web is complex, open-ended, and constantly changing, making it challenging to scale training data for visual web agents. Existing data collection attempts remain limited to offline trajectories for supervised fine-tuning or a handful of simulated environments for RL training, thus failing to capture web diversity. We propose Weblica (Web Replica), a framework for constructing reproducible and scalable web environments. Our framework leverages 1) HTTP-level caching to capture and replay stable visual states while preserving interactive behavior and 2) LLM-based environment synthesis grounded in real-world websites and core web navigation skills. Using this framework, we scale RL training to thousands of diverse environments and tasks. Our best model, Weblica-8B, outperforms open-weight baselines of similar size across multiple web navigation benchmarks while using fewer inference steps, scales favorably with additional test-time compute, and is competitive with API models.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Learning Agent Routing From Early Experience
arXiv:2605.07180v1 Announce Type: new
Abstract: LLM agents achieve strong performance on complex reasoning tasks but incur high latency and compute cost. In practice, many queries fall within the capability boundary of cutting-edge LLMs and do not require full agent execution, making effective routing between LLMs and agents a key challenge. We study the problem of routing queries between lightweight LLM inference and full agent execution under realistic cold-start settings. To address this, we propose BoundaryRouter, a training-free routing framework that uses early behavioral experience and rubric-guided reasoning to decide whether to answer a query with direct LLM inference or escalate to an agent. BoundaryRouter builds a compact experience memory by executing both systems on a shared seed set and retrieves similar cases at inference time to guide routing decisions. To evaluate this method, we introduce RouteBench, a benchmark covering in-domain, paraphrased, and out-of-domain route settings. Experiments show that BoundaryRouter reduces inference time by 60.6% compared to the agent while improving performance by 28.6% over direct LLM inference, outperforming prompt-based and retrieval-only routing by an average of 37.9% and 8.2%, respectively.
Fonte: arXiv cs.CL
Privacy/Security/Fairness • Score 85
Less Random, More Private: What is the Optimal Subsampling Scheme for DP-SGD?
arXiv:2605.07072v1 Announce Type: cross
Abstract: Poisson subsampling is the default sampling scheme in differentially private machine learning, largely because its unstructured randomness yields tractable privacy amplification analyses. Yet this same randomness introduces substantial participation variance: each sample appears in very different numbers of training iterations. In this work, we show that this variance is not merely a practical artifact to be tolerated, but a fundamental source of suboptimal privacy amplification. We prove that Balanced Iteration Subsampling (BIS), a structured scheme in which each sample participates in exactly a fixed number of iterations, achieves stronger privacy amplification than Poisson subsampling and is optimal at both extremes of the noise spectrum ($\sigma \to 0$ and $\sigma \to \infty$). Our analysis reveals that the privacy-noise tradeoff is governed not by maximizing randomness, but by eliminating participation variance while preserving uniform marginal participation across iterations. To translate this asymptotic theory into finite-noise guarantees, we introduce a practical near-exact Monte Carlo accountant for BIS, which removes the analytical slack of existing RDP and composition-based PLD analyses. Evaluations across more than 60 practical DP-SGD configurations show that BIS consistently outperforms Poisson subsampling in the low-noise regimes most relevant for high-utility private training, reducing the required noise multiplier by up to $9.6\%$. These results overturn the common intuition that more sampling randomness necessarily yields stronger privacy amplification: in DP-SGD, structured participation can be both more practical and more private. Our implementation is available at https://github.com/dong-xin-ao-andy/bis-mc-accountant.
Fonte: arXiv stat.ML
Vision • Score 85
Hard to Read, Easy to Jailbreak: How Visual Degradation Bypasses MLLM Safety Alignment
arXiv:2605.07250v1 Announce Type: new
Abstract: Recent advancements in visual context compression enable MLLMs to process ultra-long contexts efficiently by rendering text into images. However, we identify a critical vulnerability inherent to this paradigm: lowering image resolution inadvertently catalyzes jailbreaking. Our experiments reveal that the safety defenses of SOTA models deteriorate sharply as resolution degrades, surprisingly persisting even when text remains legible. We attribute this to ``Cognitive Overload'', hypothesizing that the effort required to decipher degraded inputs diverts attentional resources from safety auditing. This phenomenon is consistent across various visual perturbations, including noise and geometric distortion. To address this, we propose a simple ``Structured Cognitive Offloading'' strategy that mitigates these risks by enforcing a serialized pipeline to decouple visual transcription from safety assessment. Our work exposes a significant risk in vision-based compression and provides critical insights for the secure design of future MLLMs.
Fonte: arXiv cs.CV
Vision • Score 85
HumanNet: Scaling Human-centric Video Learning to One Million Hours
arXiv:2605.06747v1 Announce Type: new
Abstract: Progress in embodied intelligence increasingly depends on scalable data infrastructure. While vision and language have scaled with internet corpora, learning physical interaction remains constrained by the lack of large, diverse, and richly annotated human activity data. We present HumanNet, a one-million-hour human-centric video corpus that captures how humans interact with the physical world at scale. HumanNet spans both first-person and third-person perspectives and covers fine-grained activities, human-object interactions, tool use, and long-horizon behaviors across diverse real-world environments. Beyond raw video, the dataset provides interaction-centric annotations, including captions, motion descriptions, and hand and body-related signals, enabling motion-aware and interaction-aware learning. Beyond scale, HumanNet introduces a systematic data curation paradigm for embodied learning, where human-centric filtering, temporal structuring, viewpoint diversity, and annotation enrichment are treated as first-class design principles. This design transforms unstructured internet video into a scalable substrate for representation learning, activity understanding, motion generation, and human-to-robot transfer. We conduct a first-step validation on the value of this design through controlled vision-language-action ablation: under a fixed set of validation data, continued training from the Qwen VLM model with 1000 hours of egocentric video drawn from HumanNet surpasses the continued training with 100 hours of real-robot data from Magic Cobot, indicating that egocentric human video could be a scalable and cost-effective substitute for robot data. By building this project, we aim to explore the opportunity to scale embodied foundation models using human-centric videos, rather than relying solely on robot-specific data.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
PaT: Planning-after-Trial for Efficient Test-Time Code Generation
arXiv:2605.07248v1 Announce Type: new
Abstract: Beyond training-time optimization, scaling test-time computation has emerged as a key paradigm to extend the reasoning capabilities of Large Language Models (LLMs). However, most existing methods adopt a rigid Planning-before-Trial (PbT) policy, which inefficiently allocates test-time compute by incurring planning overhead even on directly solvable problems. We propose Planning-after-Trial (PaT), an adaptive policy for code generation that invokes a planner only upon verification failure. This adaptive policy naturally enables a heterogeneous model configuration: a cost-efficient model handles generation attempts, while a powerful model is reserved for targeted planning interventions. Empirically, across multiple benchmarks and model families, our approach significantly advances the cost-performance Pareto frontier. Notably, our heterogeneous configuration achieves performance comparable to a large homogeneous model while reducing inference cost by approximately 69\%.
Fonte: arXiv cs.CL
RL • Score 85
How to Compress KV Cache in RL Post-Training? Shadow Mask Distillation for Memory-Efficient Alignment
arXiv:2605.06850v1 Announce Type: new
Abstract: Reinforcement Learning (RL) has emerged as a crucial paradigm for unlocking the advanced reasoning capabilities of Large Language Models (LLMs), encompassing frameworks like RLHF and RLAIF. Regardless of the specific optimization algorithm (e.g., PPO, GRPO, or Online DPO), online RL inherently requires an exploratory trajectory generation (rollout) phase. However, for long-context reasoning tasks, this rollout phase imposes a severe ``memory wall'' due to the exorbitant Key-Value (KV) cache footprint. While applying KV cache compression during rollouts mitigates this memory overhead, it induces a critical off-policy bias. Although modern KV compression is often nearly lossless during standard inference, even minuscule approximation errors are drastically amplified by the inherent instability of RL optimization. Specifically, the sampler generates responses under a sparse context, whereas the learner updates parameters using the full, dense context. Existing statistical solutions, such as importance reweighting, struggle to correct this magnified bias, suffering from high gradient variance and severe sample inefficiency.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Securing Computer-Use Agents: A Unified Architecture-Lifecycle Framework for Deployment-Grounded Reliability
arXiv:2605.07110v1 Announce Type: new
Abstract: Computer-use agents(CUAs)are moving frombounded benchmarks toward real software environments, wherethey operate browsers, desktops, mobile applications, flesystems,terminals, and tool backends. In such settings, reliability isno longer captured by task success alone: perception errors,planning drift, memory use, tool mediation, permission scope,and runtime oversight jointly determine whether agent actionsremain aligned with user intent, Existing surveys organize theCUA landscape by methods, platforms, benchmarks, or securitythreats, but less explicitly connect capability formation, author-ity exposure, failure manifestation, and control placement. Toaddress this gap, the article develops an architecture-lifecycleframework for deployment-grounded reliability in CUAs. Thearchitectural view analyzes Perception, Decision, and Executionas coupled layers that transform software observations intoauthority-bearing actions, The lifecycle view examines Creation.Deployment, Operation, and Maintenance as stages in which priorsare learned, tools and permissions are bound, runtime trajecto.ries are stressed, and assurance must be preserved under drift.Using this lens, the analysis synthesizes representative systems,benchmarks, and security/privacy studies; distinguishes wherefailures become visible from where their enabling conditions areintroduced, and maps recurring intervention surfaces for controloversight, and assurance. OpenClaw is used only as a public moti.vating example of an open deployment pattern, not as a verifedinternal case study. The conclusion highlights open challengesin controllable grounding, long-horizon constraint preservation,safe authority binding, mixed-trust runtime defense, privacy-preserving memory,and continual assurance.
Fonte: arXiv cs.CL
Theory/Optimization • Score 85
Ensemble Distributionally Robust Bayesian Optimisation
arXiv:2605.07565v1 Announce Type: cross
Abstract: We study zeroth-order optimisation under context distributional uncertainty, a setting commonly tackled using Bayesian optimisation (BO). A prevailing strategy to make BO more robust to the complex and noisy nature of data is to employ an ensemble as the surrogate model, thereby mitigating the weaknesses of any single model. In this study, we propose a novel algorithm for Ensemble Distributionally Robust Bayesian Optimisation that remains computationally tractable while managing continuous context. We obtain theoretical sublinear regret bounds, improving current state-of-the-art results. We show that our method's empirical behaviour aligns with its theoretical guarantees.
Fonte: arXiv stat.ML
MLOps/Systems • Score 85
Resource-Element Energy Difference for Noncoherent Over-the-Air Federated Learning
arXiv:2605.07263v1 Announce Type: cross
Abstract: Over-the-air federated learning (OTA-FL) reduces uplink latency by exploiting waveform superposition, but conventional analog aggregation schemes typically require instantaneous channel state information (CSI), channel inversion, and coherent phase alignment, which can be difficult to maintain in practical wireless systems. This paper proposes resource-element energy difference (REED), a noncoherent aggregation primitive for continuous signed updates that avoids instantaneous CSI. REED maps the positive and negative parts of each real-valued update to transmit energies on two orthogonal resource elements with independent phase dithers, and the server estimates the signed aggregate from their energy difference. With only slow-timescale calibration of average channel powers, REED is unbiased for the desired signed sum and admits an exact closed-form variance under Rayleigh fading. We incorporate REED into full-participation FedAvg and prove a smooth nonconvex stationarity bound. Under an average per-client energy budget, the aggregation gain can be scheduled so that the REED-induced perturbation scales quadratically with the local stepsize, yielding the canonical (1/sqrt(T)) stationarity rate. Experiments on MNIST and Fashion-MNIST demonstrate that REED closely matches clean FedAvg and coherent CSIT aggregation in IID settings, while maintaining stable convergence with a moderate performance degradation under strong data heterogeneity.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
SHARP: A Self-Evolving Human-Auditable Rubric Policy for Financial Trading Agents
arXiv:2605.06822v1 Announce Type: new
Abstract: Large language models (LLMs) are increasingly deployed for autonomous financial trading, a domain requiring continuous adaptation to noisy, non-stationary markets. Existing self-improving agents typically address this through unbounded free-form prompt optimization. However, in low signal-to-noise environments with delayed scalar rewards (P\&L), this unstructured approach exacerbates the fundamental credit assignment problem: optimizers cannot reliably distinguish systematic logic flaws from stochastic market variance, inevitably leading to policy drift. To overcome this bottleneck, we introduce the Self-Evolving Human-Auditable Rubric Policy (SHARP), a neuro-symbolic framework that replaces unconstrained text mutation with structured, symbolic policy optimization. SHARP confines the agent's reasoning to a bounded, human-readable rubric of explicit condition-action rules. When sub-optimal trades occur, an attribution agent employs cross-sample reasoning across multiple samples to isolate specific rule failures. This enables targeted, atomic policy edits that are subsequently regularized through strict walk-forward validation. Evaluated across three diverse equity sectors and four LLM backbones, SHARP consistently transforms generic initial heuristics into highly robust strategies, lifting the empirical performance of compact models by 10 to 20 percentage points on average (e.g., GPT-4o-mini). Ultimately, SHARP demonstrates that LLMs can achieve dynamic and efficient adaptation while significantly enhancing the structural transparency and auditability demanded by institutional finance.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
TopoFisher: Learning Topological Summary Statistics by Maximizing Fisher Information
arXiv:2605.07720v1 Announce Type: new
Abstract: Persistence diagrams provide stable, interpretable summaries of geometric and topological structure and are useful for simulation-based inference when low-order statistics miss key information. Yet persistence-based pipelines require hand-chosen filtrations, vectorizations, and compressors, typically without an objective tied to parameter uncertainty. We introduce \textbf{TopoFisher}, a differentiable persistent-homology pipeline that learns topological summaries by maximizing local Gaussian Fisher information. Using simulations near a fiducial parameter, TopoFisher optimizes trainable filtrations, diagram vectorizations, and compressors without posterior samples or supervised regression targets, while retaining stable topological inductive bias. We also give sufficient regularity conditions for the log-determinant Fisher loss to be locally Lipschitz in trainable parameters. Controlled experiments on noisy spirals and Gaussian random fields, where total Fisher information is known, show that TopoFisher recovers much of the available information and outperforms fixed topological vectorizations. Our main results are on weak gravitational lensing, a high-dimensional non-Gaussian cosmological field-inference problem. Learned topological summaries reach higher Fisher information than state-of-the-art cosmological summaries and approach an unconstrained Information Maximising Neural Network baseline with up to $\sim80\times$ fewer parameters. The learned filtrations also generalize better: under simulator shift from lognormal to LPT-based maps it retains most Fisher information, while the neural baseline drops, and in neural posterior estimation they give tighter constraints than the neural baseline, and of state-of-the-art cosmological summaries. These results support Fisher-based topological optimization as a robust, parameter-efficient front end for simulation-based inference.
Fonte: arXiv stat.ML
Theory/Optimization • Score 85
Christoffel-DPS: Optimal sensor placement in diffusion posterior sampling for arbitrary distributions
arXiv:2605.06861v1 Announce Type: new
Abstract: State estimation is a critical task in scientific, engineering and control applications. Since the reliability of reconstructions depends on the number and position of sensors, optimal sensor placement (OSP) is essential in scenarios where measurements are sparse and expensive. Classical OSP approaches rely on Gaussian assumptions and are consequently unable to account for the complex distributions encountered in many real-world systems. Generative-model-based reconstruction using sensor guided diffusion posterior sampling (DPS) has emerged as a promising technique for reconstructing states from highly complex distributions. However, existing sensor-selection methods either require unrealistically many sensors or emulate classical OSP, creating a mismatch between modern recovery models with classical OSP tools motivating the need for fundamentally new ideas towards OSP that match the recent advances made in powerful recovery models. We introduce a distribution-free sensor placement framework based on the Christoffel function: a mathematical formulation of optimal sampling and recovery guarantees for posterior sampling with arbitrary sensors and signal distributions, from which we derive a new OSP strategy with non-asymptotic bounds on the number of sensors needed for recovery. We develop Christoffel-DPS, with offline and online variants, instantiating Christoffel sampling for generative models. Christoffel-DPS outperforms Gaussian OSP baselines and existing generative-model placement methods, validating that distribution-free sensing is both theoretically principled and practically superior. The framework is model-agnostic; we demonstrate its application to a range of unconditional DPS and flow-matching models on structurally non-Gaussian benchmarks, showing the efficacy of Christoffel-DPS in low sensor budget regimes.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Physics-based Digital Twins for Integrated Thermal Energy Systems Using Active Learning
arXiv:2605.06756v1 Announce Type: new
Abstract: Real-time supervisory control of thermal energy distribution systems requires digital twins that are accurate, interpretable, and uncertainty-aware, yet remain data and computationally efficient. High-fidelity simulations alone are costly, while purely data-driven surrogates often lack robustness. To address these challenges, this work proposes an active learning (AL) framework that couples system-level Modelica simulations with four simpler physics-informed and data-driven surrogate modeling approaches: deterministic Sparse Identification of Nonlinear Dynamics with Control (SINDyC), its probabilistic multivariate-Gaussian extension (MvG-SINDyC), feedforward neural network (FNN), and gated recurrent unit (GRU) network. Tailored to each surrogate, model-specific AL query strategies are employed, including Mahalanobis-distance sampling in coefficient space for MvG-SINDyC and error-based sampling in prediction space for SINDyC, FNN, and GRU, allowing the learning process to prioritize dynamically informative trajectories.
The proposed approach is demonstrated on the glycol heat exchanger (GHX) subsystem of the Thermal Energy Distribution System (TEDS) at Idaho National Laboratory. Across key GHX outputs--the bypass mass flow rate $\dot{m}_{\mathrm{GHX}}$ and heat transfer rate $Q_{\mathrm{GHX}}$-the AL framework achieves comparable predictive accuracy using as few as one-fifth of the simulation trajectories required by random sampling. Among the evaluated surrogates, the GRU achieves the highest predictive fidelity, while SINDyC remains the most computationally efficient and interpretable. The probabilistic MvG-SINDyC surrogate further enables uncertainty quantification and exhibits the largest computational gains under AL.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
The E$\Delta$-MHC-Geo Transformer: Adaptive Geodesic Operations with Guaranteed Orthogonality
arXiv:2605.06729v1 Announce Type: new
Abstract: We present the E$\Delta$-MHC-Geo Transformer, a novel architecture that unifies Manifold-Constrained Hyper-Connections (mHC), Deep Delta Learning (DDL), and the Cayley transform to obtain input-adaptive, unconditionally orthogonal residual connections. Unlike DDL, whose Householder operator is orthogonal only at $\beta \in \{0,2\}$, our Data-Dependent Cayley rotation $Q(x)=(I+(\beta/2)A(x))^{-1}(I-(\beta/2)A(x))$ preserves orthogonality for all $\beta$ and all inputs. To handle negation, an eigenvalue $-1$ case that Cayley provably excludes, we introduce the E$\Delta$-MHC-Geo Hybrid, which combines Cayley rotation with Householder reflection via a learned operator-selection gate $X'=\gamma(X)Q(X)X+(1-\gamma(X))H_2(X)X$. A midpoint-collapse regularizer, $4\gamma(1-\gamma)$, encourages boundary gate decisions, where each selected component is orthogonal. In matched-parameter comparisons, with approximately 1.79M parameters per model and mean +/- standard deviation over 3 seeds, against four baselines including the concurrent JPmHC, E$\Delta$-MHC-Geo achieves the best long-horizon stability, 1.9x over JPmHC and 3.8x over GPT; the best near-$\pi$ rotation loss, 4.5x over JPmHC on single-plane; strong norm preservation, with 0.001 mean deviation; and 0.96 negation cosine alignment in a diagnostic reflection probe, all with 33% fewer layers. While JPmHC's wider representation excels on pure rotation, its finite Cayley residual mixer excludes an exact $\lambda=-1$ operator and has no reflection branch, motivating our hybrid approach for accessing both connected components of $O(n)$.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Sparse Attention as a Range Searching Problem: Towards an Inference-Efficient Index for KV Cache
arXiv:2605.06763v1 Announce Type: new
Abstract: Sparse attention improves LLM inference efficiency by selecting a subset of key-value entries, but at the cost of potential accuracy degradation. In particular, omitting critical KV entries can induce substantial errors in model outputs. Existing methods typically operate under fixed or adaptive token budgets and provide empirical robustness or partial theoretical guarantees, yet they do not ensure zero false negatives in decoding steps, particularly since the set of relevant tokens is both query- and step-dependent. Our empirical observations confirm that missing even one critical key can lead to sharp error spikes, especially in long reasoning tasks where the set of important tokens varies throughout decoding. This observation motivates the need for indexing methods that dynamically adapt to these variations across decoding steps while guaranteeing a full recall of the relevant keys above a certain threshold. We address this challenge by reformulating sparse attention as the halfspace range searching problem. However, existing range searching indices are not suitable for modern LLM inference due to their computational and implementation overheads. To overcome this, we introduce Louver, a novel index structure tailored for efficient KV cache retrieval. Louver (i) guarantees zero false negatives with respect to a specified threshold in both theory and practice, (ii) is lightweight to integrate into existing LLM pipelines, and (iii) incorporates hardware-aware optimizations for both CPU and GPU executions. Our experiments demonstrate that Louver outperforms prior sparse attention methods in both accuracy and runtime, and is faster than highly optimized dense attentions such as FlashAttention. These results highlight that recall guarantees are a critical and overlooked dimension of sparse attention, and open a new direction for building theoretically grounded, efficient KV cache indices.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
SREGym: A Live Benchmark for AI SRE Agents with High-Fidelity Failure Scenarios
arXiv:2605.07161v1 Announce Type: new
Abstract: AI agents are increasingly used to diagnose and mitigate failures in production systems, known as agentic Site Reliability Engineering (SRE). Current SRE benchmarks are limited to oversimplistic SRE tasks and are unfortunately hard to extend due to bespoke designs. We present SREGym, a high-fidelity benchmark for SRE agents. SREGym exposes a live system environment built atop real-world cloud-native system stacks, where high-fidelity failure scenarios are simulated through fault injectors. SREGym models the complexity of production environments by simulating (1) a wide range of faults at different layers, (2) various ambient noises, and (3) diverse failure modes such as metastable failures and correlated failures. SREGym is architected as a modular, extensible framework that orchestrates fault and noise injectors across stacks. SREGym currently includes 90 realistic, challenging SRE problems. We use SREGym to evaluate frontier agents and show that their capabilities varies significantly in addressing different kinds of failures, with up to 40% differences in end-to-end results. SREGym is actively maintained as an open-source project and has been used by researchers and practitioners.
Fonte: arXiv cs.AI
MLOps/Systems • Score 90
Switchcraft: AI Model Router for Agentic Tool Calling
arXiv:2605.07112v1 Announce Type: new
Abstract: Agentic AI systems that invoke external tools are powerful but costly, leading developers to default to large models and overspend inference budgets. Model routing can mitigate this, but existing routers are designed for chat completion rather than tool use. We present Switchcraft, the first (to the best of our knowledge) model router optimized for agentic tool calling. Switchcraft operates inline, selecting the lowest-cost model subject to correctness. We construct an evaluation framework on five function-calling benchmarks and train a DistilBERT-based classifier, deployed under a latency budget. Switchcraft achieves 82.9% accuracy -- matching or exceeding the best individual model -- while reducing inference cost by 84%, saving over $3,600 per million queries. We find that larger models do not consistently outperform smaller ones on tool-use tasks, and that nominally cheaper models can incur higher total cost due to token-intensive reasoning. Our work enables cost-aware agentic AI deployment without sacrificing correctness.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Towards Autonomous Business Intelligence via Data-to-Insight Discovery Agent
arXiv:2605.07202v1 Announce Type: new
Abstract: Transforming fragmented enterprise data into actionable insights remains a significant challenge for LLMs, constrained by complex database schemas, limitations in dynamic SQL generation, and the need for deep multi-dimensional analysis.In this paper, we propose AIDA(Autonomous Insight Discovery Agent), the first end-to-end framework designed for autonomous exploration in complex business environments. We establish a highly flexible instant retail environment encompassing 200+ metrics and 100+ dimensions, and integrates a proprietary Domain-Specific Language (DSL) that bridges semantic reasoning with precise SQL execution. Our reinforcement learning system subsequently formulates business analysis as a Pareto Principle-guided cumulative reasoning process. Experimental results demonstrate that AIDA significantly outperforms workflow-based agents, and extensive evaluations further reveal that AIDA achieves superior environmental perception and more in-depth analysis from diverse perspectives. Our work ultimately establishes the transformative potential of autonomous intelligence for industrial-scale business intelligence systems.
Fonte: arXiv cs.AI
NLP/LLMs • Score 92
2.5-D Decomposition for LLM-Based Spatial Construction
arXiv:2605.07066v1 Announce Type: new
Abstract: Autonomous systems that build structures from natural-language instructions need reliable spatial reasoning, yet large language models (LLMs) make systematic coordinate errors when generating three-dimensional block placements. We present a neuro-symbolic pipeline based on \emph{2.5-D decomposition}: the LLM plans in the two-dimensional horizontal plane while a deterministic executor computes all vertical placement from column occupancy, eliminating an entire class of errors. On the Build What I Mean benchmark (160 rounds), GPT-4o-mini with this pipeline achieves 94.6\% mean structural accuracy across 12 independent runs, within 3.0 percentage points of the 97.6\% ceiling imposed by architect-agent errors that no builder-side improvement can address. This outperforms both GPT-4o at 90.3\% and the best competing system at 76.3\%. A controlled ablation confirms that 2.5-D decomposition is the dominant contributor, accounting for 50.7 percentage points of accuracy. The pipeline transfers directly to edge hardware: Nemotron-3 120B running locally on an NVIDIA Jetson Thor AGX matches the cloud result at 94.5\% with no prompt modifications. The underlying principle, removing deterministic dimensions from the LLM's output space, applies to any autonomous construction or assembly task where gravity or other physical constraints fix one or more degrees of freedom. A transfer experiment on 500 IGLU collaborative building tasks confirm the effect generalizes beyond the primary benchmark.
Fonte: arXiv cs.AI
RL • Score 85
Randomness is sometimes necessary for coordination
arXiv:2605.06825v1 Announce Type: new
Abstract: Full parameter sharing is standard in cooperative multi-agent reinforcement learning (MARL) for homogeneous agents. Under permutation-symmetric observations, however, a shared deterministic policy outputs identical action distributions for every agent, making role differentiation impossible. This failure can theoretically be resolved using symmetry breaking among anonymous identical processors, which requires randomness. We propose Diamond Attention, a cross-attention architecture in which each agent samples a scalar random number per timestep, inducing a transient rank ordering that masks lower-ranked peers from agent-to-agent attention while leaving task attention fully unmasked. This realizes a random-bit coordination protocol in a single broadcast round, and the set-based attention enables zero-shot deployment to teams of different sizes. We evaluate across three regimes that isolate when structured randomness matters. On the perfectly symmetric XOR game, our method achieves $1.0$ success while all deterministic baselines plateau near $0.5$. On control coordination tasks, a policy trained on $N=4$ generalizes zero-shot to $N \in [2,8]$. On SMACLite cross-scenario transfer, we achieve zero-shot transfer where standard baselines cannot transfer due to structural limitations. Furthermore, replacing the structured mask with standard dropout-based randomness results in a 0\% win rate, confirming that protocol-space structure, not stochastic noise, is the operative ingredient. https://anonymous.4open.science/r/randomness-137A/
Fonte: arXiv cs.AI
Applications • Score 85
Hierarchical Perfusion Graphs for Tumor Heterogeneity Modeling in Glioma Molecular Subtyping
arXiv:2605.07156v1 Announce Type: new
Abstract: Precise molecular subtyping of gliomas, including isocitrate dehydrogenase (IDH) mutation and 1p/19q codeletion, directly guides surgical and therapeutic decisions, yet currently relies on invasive tissue sampling. Deep learning on structural MRI has emerged as a non-invasive alternative, but anatomy-only approaches cannot capture the hemodynamic signatures that distinguish molecular subtypes. Radiogenomics based on dynamic susceptibility contrast (DSC) MRI holds immense potential for non-invasively characterizing glioma molecular subtypes, yet clinical deployment has been hindered by inter-site variability and the limitations of voxel-wise analysis. We introduce HiPerfGNN, a framework that first learns discrete hemodynamic representations from raw time-intensity curves using a vector-quantized variational autoencoder (VQ-VAE). These quantized perfusion codes define coarse-level graph nodes representing functional tumor habitats, each of which is hierarchically subdivided into fine-level subregions guided by structural MRI. A hierarchical graph neural network then propagates information across scales for molecular prediction. On an internal cohort (n=475), the model achieved AUCs of 0.96 (IDH), 0.89 (1p/19q), and 0.84 (WHO grade), and maintained robust IDH performance (AUC 0.89) on an independent external cohort (n=397) without recalibration. Gradient-based saliency analysis confirms biologically grounded attention patterns aligned with known glioma pathophysiology. Our results demonstrate the added value of integrating perfusion dynamics into radiogenomic pipelines for glioma molecular subtyping. Code is available at https://github.com/janghana/HiPerfGNN.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
LENS: Low-Frequency Eigen Noise Shaping for Efficient Diffusion Sampling
arXiv:2605.07253v1 Announce Type: new
Abstract: Distilled diffusion models accelerate image generation by reducing the number of denoising steps, but often suffer from degraded image quality. To mitigate this trade-off, test-time optimization methods improve quality, yet their iterative nature incurs substantial computational overhead and leads to slow inference, limiting practical usability. Recent hypernetwork-based approaches amortize this process during training, but still require costly noise modulation in high-dimensional latent spaces. In this work, we propose LENS (Low-frequency Eigen Noise Shaping), an efficient noise modulation framework that operates in a low-dimensional subspace. Our approach is motivated by the observation that low-frequency components of the noise largely determine the global structure and visual fidelity of generated images. Based on this observation, we provide a theoretical justification for restricting modulation to the low-frequency subspace and derive a principled training objective. Building on this, LENS employs a lightweight, standalone network to selectively modulate these components, enabling efficient and targeted noise modulation. Extensive experiments demonstrate that LENS achieves competitive image quality while reducing FLOPs by 400-700$\times$, model parameters by 25-75$\times$, and inference-time overhead by 10-20$\times$ compared to prior methods.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
CASCADE: Context-Aware Relaxation for Speculative Image Decoding
arXiv:2605.07230v1 Announce Type: new
Abstract: Autoregressive generation is a powerful approach for high-fidelity image synthesis, but it remains computationally demanding and slow even on the most advanced accelerators. While speculative decoding has been explored to mitigate this bottleneck, existing approaches fail to achieve efficiency gains comparable to those observed in text generation. A key limitation is the target model's high uncertainty during image generation, which leads to high draft token rejection rates. In this work, we identify previously overlooked patterns in the target model's behavior that emerge naturally in tree-based speculative decoding. Specifically, we formalize two properties, semantic interchangeability and convergence, arising from the redundancies in the target model's hidden state representations. By capturing these redundancies across the depth and breadth of the predicted token tree, our method identifies principled opportunities for acceptance relaxation without requiring additional training. Additionally, we enhance standalone drafter performance by injecting the redundancy signals from the target model into drafter training with minimal modification. We evaluate our approach across multiple text-to-image models and drafter architectures. Results show that CASCADE achieves state-of-the-art speedups for drafter-based speculative decoding, with up to 3.6x acceleration, while maintaining image quality and text-prompt fidelity.
Fonte: arXiv cs.CV
Vision • Score 85
Closed-Form Linear-Probe Dataset Distillation for Pre-trained Vision Models
arXiv:2605.07194v1 Announce Type: new
Abstract: Dataset distillation compresses a large training set into a small synthetic set that preserves downstream training utility. While most existing methods target training networks from scratch, modern visual transfer learning often uses frozen pre-trained encoders followed by lightweight linear probing. Existing distillation methods for this setting either unroll iterative linear-probe updates with trajectory-based gradient matching, or rely on closed-form formulations originally designed for from-scratch training with neural-tangent-kernel (NTK) approximations. Neither route exploits the fact that frozen-feature linear probing admits a closed-form solution determined directly by the pre-trained features themselves, with no infinite-width approximation and no inner-loop trajectory. We propose Closed-Form Linear-Probe Dataset Distillation (CLP-DD), a bilevel formulation that computes the linear probe induced by the synthetic set with a sample-space kernel ridge solver. The synthetic images are then updated by evaluating this induced classifier on real features through a temperature-scaled softmax cross-entropy, where the classifier columns act as learned class anchors in feature space. We further show that the choice of outer objective is decisive: pairing the closed-form inner solver with a standard MSE outer loss substantially underperforms trajectory-based methods, while the discriminative outer loss closes most of the gap. On ImageNet-100 with four pre-trained backbones, CLP-DD substantially improves over LGM without DSA and approaches LGM with DSA at a fraction of the computational cost. On ImageNet-1K, CLP-DD matches or surpasses LGM with DSA on three of four backbones while running roughly $14\times$ faster and using less than one-eighth of the GPU memory.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Reformulating KV Cache Eviction Problem for Long-Context LLM Inference
arXiv:2605.07234v1 Announce Type: new
Abstract: Large language models (LLMs) support long-context inference but suffer from substantial memory and runtime overhead due to Key-Value (KV) Cache growth. Existing KV Cache eviction methods primarily rely on local attention weights, neglecting the influence of value representations, output projection, and inter-head interactions. In this work, we reformulate KV Cache eviction from a conventional head-wise, weight-averaging approach into an output-aware, layer-wise matrix multiplication approximation problem. We introduce LaProx, a novel eviction strategy that explicitly models the multiplicative interaction between attention maps and projected value states to accurately quantify token contributions while accounting for inter-head dependencies. Building on this metric, we propose the first unified eviction strategy that assigns globally comparable importance scores to tokens, enabling model-wide selection instead of local, head-wise decisions. Experimental results across 19 datasets on long-context benchmarks LongBench and Needle-In-A-Haystack demonstrate that our approach maintains model performance with only 5\% of the KV cache and consistently outperforms prior works across all configurations. Notably, our method achieves up to 2$\times$ accuracy loss reduction under extreme compression scenarios compared to existing state-of-the-art baselines with minimal overhead.
Fonte: arXiv cs.CL
Vision • Score 85
XiYOLO: Energy-Aware Object Detection via Iterative Architecture Search and Scaling
arXiv:2605.06927v1 Announce Type: new
Abstract: Object detection on heterogeneous edge devices must satisfy strict energy, latency, and memory constraints while still providing reliable perception for downstream autonomy. Existing energy-aware NAS methods often target limited deployment settings, while real energy remains difficult to optimize because it is highly device-dependent and costly to measure. We address these challenges with an energy-adaptive framework that combines an energy-aware XiResOFA search space, a two-stage energy estimator, and iterative search to identify a single energy-efficient base architecture. We then apply compound scaling to transform this base design into the XiYOLO family across deployment budgets, enabling interpretable accuracy-energy tradeoffs under sparse hardware measurements. Experiments on PascalVOC, COCO, and real-device deployment show that XiYOLO achieves a stronger energy-accuracy tradeoff than YOLO baselines. On PascalVOC, the medium XiYOLO model reaches 86.15 mAP50 while reducing energy relative to YOLOv12m by 20.6% on GPU and 35.9% on NPU. On COCO, XiYOLO reduces energy relative to YOLOv12 by up to 53.7% on GPU and 51.6% on NPU at the small scale. The proposed two-stage estimator also improves sample efficiency over a joint predictor under few-shot adaptation with only 2-20 target-device samples.
Fonte: arXiv cs.CV
Theory/Optimization • Score 85
A Rod Flow Model for Adam at the Edge of Stability
arXiv:2605.06821v1 Announce Type: new
Abstract: Cohen et al. (arXiv:2207.14484) observed that adaptive gradient methods such as Adam operate at the edge of stability. While there has been significant work on continuous-time modeling of gradient descent at the edge of stability, extending these models to momentum methods remains underdeveloped. In the gradient descent setting, Regis et al. (arXiv:2602.01480) introduced rod flow, which models consecutive iterates as an extended one-dimensional object -- a "rod." Here we extend rod flow to Adam by working in the joint phase space of parameters and first moment $(w, m)$ and treating the second moment $\nu$ as a smooth auxiliary variable. We also develop rod flows for heavy ball momentum, Nesterov momentum, and scalar and per-component versions of RMSProp, Adam, and NAdam. For all eight optimizers, we empirically evaluate rod flow on representative machine learning architectures, where it tracks the discrete iterates through the edge-of-stability regime significantly more accurately than the corresponding stable flow.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
CLIPer: Tailoring Diverse User Preference via Classifier-Guided Inference-Time Personalization
arXiv:2605.07162v1 Announce Type: new
Abstract: Personalized LLMs can significantly enhance user experiences by tailoring responses to preferences such as helpfulness, conciseness, and humor. However, fine-tuning models to address all possible combinations of user preferences is computationally expensive and impractical. In this paper, we introduce \textbf{CLIPer}(\textbf{Cl}assifier-guided \textbf{I}nference-time \textbf{Per}sonalization), a lightweight personalization approach that leverages a classifier model to steer LLM generation dynamically to different user preferences at inference time. Our method eliminates the need for extensive fine-tuning, inducing negligible additional computational overhead while enabling more controllable and nuanced personalization across single and multi-dimensional preferences. Comprehensive empirical analyses demonstrate the scalability and effectiveness of our approach in delivering personalized language generation.
Fonte: arXiv cs.CL
Privacy/Security/Fairness • Score 85
Modulated learning for private and distributed regression with just a single sample per client device
arXiv:2605.07233v1 Announce Type: cross
Abstract: This work focuses on the question of learning from a large number of devices with each device holding only a single sample of data. Several real-world applications exist to this one sample per client setup up including learning from fitness trackers, data/app usage aggregators, body-worn sensing devices, and daily event monitors to name a few. When a client has only one sample, the standard federated learning paradigm breaks down as a local update based on that single point is far from being useful, especially in the earlier rounds for estimation of the model coefficients. This utility is further weakened by the privacy-inducing noise applied at every round. This work caters to this problem to enable such clients to collaboratively contribute to effectively learn a global model without leaking the privacy of their data. The proposed approach injects a single, carefully calibrated noisy perturbation to transform the sample at each client, followed by a post-processed representation which is shared with the server. These representations aggregated at the server are processed to obtain an unbiased gradient update that in expectation matches the non-private centralized gradient while preserving data privacy. This approach is different than traditional private federated learning, where the communication payloads involve model coefficients as opposed to privately transformed data samples. This method enables devices with extremely limited data to collaborate and learn accurate, privacy-preserving models without requiring large local datasets or sacrificing individual privacy.
Fonte: arXiv stat.ML
Theory/Optimization • Score 85
Open-Ended Task Discovery via Bayesian Optimization
arXiv:2605.07572v1 Announce Type: cross
Abstract: When applying Bayesian optimization (BO) to scientific workflow, a major yet often overlooked source of uncertainty is the task itself -- namely, what to optimize and how to evaluate it -- which can evolve as evidence accumulates. We introduce Generate-Select-Refine (GSR), a open-ended BO framework that alternates between task generation and task optimization. Starting from a user-provided seed task, GSR generates new tasks in a coarse-to-fine manner while a task-acquisition function schedules optimization. Asymptotically, it concentrates evaluations on the best task, incurring only logarithmic regret overhead relative to single-task BO. We apply GSR to new product development, chemical synthesis scaling, algorithm analysis, and patent repurposing, where it outperforms existing LLM-based optimizers.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Not All Tokens Need 40 Steps: Heterogeneous Step Allocation in Diffusion Transformers for Efficient Video Generation
arXiv:2605.06892v1 Announce Type: new
Abstract: Diffusion Transformers (DiTs) have achieved state-of-the-art video generation quality, but they incur immense computational cost because standard inference applies the same number of denoising steps uniformly to every token in the sequence. It is well known that human vision ignores vast amounts of redundant motion. Why, then, do our densest models treat every spatiotemporal token with equal priority? In this paper, we introduce Heterogeneous Step Allocation (HSA), a training-free inference algorithm that assigns varying step budgets to different spatiotemporal tokens based on their velocity dynamics. To resolve the resulting sequence-length mismatch without sacrificing global context, HSA introduces a KV-cache synchronization mechanism that allows active tokens to attend to the full sequence while entirely bypassing inactive tokens. Furthermore, we derive a cached Euler update that advances the latent states of skipped tokens in a single operation without additional model evaluations. We evaluate HSA on the Wan-2 and LTX-2 models for both text-to-video (T2V) and image-to-video (I2V) generation. Our results demonstrate that HSA significantly outperforms previous state-of-the-art caching methods and the vanilla Flow Matching baseline, especially at aggressive acceleration regimes (e.g., 50% and 25% runtimes). Crucially, HSA achieves a superior quality-runtime Pareto frontier without the need for expensive offline profiling, robustly preserving structural integrity and generation quality even under tight computational budgets.
Project page: https://ernestchu.github.io/hsa
Fonte: arXiv cs.CV
RL • Score 85
Agentick: A Unified Benchmark for General Sequential Decision-Making Agents
arXiv:2605.06869v1 Announce Type: new
Abstract: AI agent research spans a wide spectrum: from RL agents that learn from scratch to foundation model agents that leverage pre-trained knowledge, yet no unified benchmark enables fair comparison across these approaches. We present Agentick, a benchmark for sequential decision-making agents designed to evaluate RL, LLM, VLM, hybrid, and human agents on common ground and to power research on the fundamental challenges of sequential decision-making. Agentick provides 37 procedurally generated tasks across six capability categories, four difficulty levels, and five observation modalities, all exposed through a single Gymnasium-compatible interface. The benchmark ships with a Coding API, oracle reference policies for all tasks, pre-built SFT datasets, a composable agent harness, and a live leaderboard. An evaluation spanning 27 configurations and over 90,000 episodes reveals that no single approach dominates: GPT-5 mini leads overall at 0.309 oracle-normalized score while PPO dominates planning and multi-agent tasks; the reasoning harness multiplies LLM performance by 3-10x; and ASCII observations consistently outperform natural language. These findings highlight the substantial room for improvement that remains across all agent paradigms. Agentick's capability-decomposed, multi-modal design provides the empirical infrastructure needed to drive progress toward general autonomous agents, both as an evaluation framework and as a training ground for RL post-training of foundation models in truly sequential environments.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
Functional-prior-based Bayesian PDE-constrained inversion using PINNs
arXiv:2605.07060v1 Announce Type: cross
Abstract: Physics-informed neural networks (PINNs) provide a mesh-free framework for solving PDE-constrained inverse problems, but their extension to Bayesian inversion still faces a fundamental difficulty: prior distributions are typically defined in the weight space of neural networks, whereas physically meaningful prior assumptions are more naturally expressed in function space. In this study, we introduce a unified framework, termed functional-prior-based approaches to Bayesian PDE-constrained inversion using physics-informed neural networks (fpBPINN), to incorporate functional priors into Bayesian PINN-based inversion. We consider two complementary approaches. The first is a functional-prior-informed Bayesian PINN (FPI-BPINN), in which a neural network weight prior is learned to be consistent with a prescribed functional prior, and Bayesian inference is subsequently performed in weight space. The second is function-space particle-based variational inference for PINNs (fParVI-PINN), which performs Bayesian estimation using ParVI directly in function space. We also show that random Fourier features (RFF) play an important role in representing Gaussian functional priors with neural networks and in improving posterior approximation. We applied the proposed approaches to one-dimensional seismic traveltime tomography and two-dimensional Darcy-flow permeability inversion. These numerical experiments showed that both approaches accurately estimated posterior distributions, highlighting the significance of introducing physically interpretable functional priors into Bayesian PINN-based inverse problems. We also identified the contrasting advantages of FPI-BPINN and fParVI-PINN, namely flexibility and accuracy, respectively.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
RateQuant: Optimal Mixed-Precision KV Cache Quantization via Rate-Distortion Theory
arXiv:2605.06675v1 Announce Type: new
Abstract: Large language models cache all previously computed key-value (KV) pairs during generation, and this KV cache grows linearly with sequence length, making it a primary memory bottleneck for serving. Quantizing the KV cache to fewer bits reduces this cost, yet all current quantizers assign the same bit-width to every attention head, ignoring the large variation in head importance. A natural idea is to allocate more bits to important heads and fewer to the rest. We show, however, that such mixed-precision allocation has a hidden pitfall: each quantizer follows a different distortion curve D(b)=alpha*beta^{-b}, and the decay rate beta varies from 3.6 to 5.3 across quantizer designs. Applying one quantizer's distortion model to another inverts the allocation order and makes performance worse than uniform quantization. We call this failure mode distortion model mismatch and propose RateQuant to resolve it. RateQuant fits a per-quantizer distortion model from a small calibration set, then solves the resulting bit-allocation problem in closed form via reverse waterfilling from rate-distortion theory. On Qwen3-8B at 2.5 average bits, calibrated RateQuant reduces KIVI's perplexity from 49.3 to 14.9 (70% reduction) and improves QuaRot by 6.6 PPL. The entire calibration takes 1.6 s on a single GPU and adds zero overhead at inference time.
Fonte: arXiv cs.LG
Privacy/Security/Fairness • Score 85
SLAM: Structural Linguistic Activation Marking for Language Models
arXiv:2605.05443v1 Announce Type: new
Abstract: LLM watermarks must be detectable without compromising text quality, yet most existing schemes bias the next-token distribution and pay for detection with measurable quality loss. We present SLAM (Structural Linguistic Activation Marking), a novel white-box watermarking scheme that sidesteps this cost by writing the mark into structural geometry rather than token frequencies: sparse autoencoders identify residual-stream directions encoding linguistic structure (e.g., voice, tense, clause order), and we causally steer those directions at generation time, leaving lexical sampling and semantics unconstrained. On Gemma-2 2B and 9B, SLAM achieves 100% detection accuracy with a quality cost of only 1-2 reward points - compared to 7.5-11.5 for KGW, EWD, and Unigram - with naturalness and diversity preserved at near-unwatermarked levels across both models. The trade-off is a complementary robustness profile: SLAM resists word-level edits but is vulnerable to paraphrase that restructures syntax (at a quality cost), the converse of token-distribution methods.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
UniPrefill: Universal Long-Context Prefill Acceleration via Block-wise Dynamic Sparsification
arXiv:2605.06221v1 Announce Type: new
Abstract: As large language models (LLMs) continue to advance rapidly, they are becoming increasingly capable while simultaneously demanding ever-longer context lengths. To improve the inference efficiency of long-context processing, several novel low-complexity hybrid architectures have recently been proposed, effectively alleviating the computational burden of long-context inference. However, existing research on long-context prefill acceleration remains predominantly focused on sparse attention mechanisms, which achieve their maximum speedup only on full-attention models. When transferred to emerging architectures--such as linear/full attention hybrids or sliding window/full attention hybrids--these prefill acceleration approaches suffer significant performance degradation. Furthermore, such methods are generally incompatible with continuous batching, making them difficult to integrate into modern inference engines such as vLLM. To this end, we propose UniPrefill, a prefill acceleration framework applicable to virtually any model architecture, which directly accelerates the model's computation at the token level. We further implement UniPrefill as a continuous batching operator and extend vLLM's scheduling strategy to natively support prefill-decode co-processing and tensor parallel for UniPrefill, enabling its seamless integration into vLLM. UniPrefill achieves up to 2.1x speedup in Time-To-First-Token (TTFT), with the acceleration becoming increasingly pronounced as the number of concurrent requests grows.
Fonte: arXiv cs.CL
MLOps/Systems • Score 85
MANTRA: Synthesizing SMT-Validated Compliance Benchmarks for Tool-Using LLM Agents
arXiv:2605.06334v1 Announce Type: new
Abstract: Tool-using large language model (LLM) agents are increasingly deployed in settings where their reliable behavior is governed by strict procedural manuals. Ensuring that such agents comply with the rules from these manuals is challenging, as they are typically written for humans in natural language while agent behavior manifests as an execution trace of tool calls. Existing evaluations of LLM agents rely on manually constructed benchmarks or LLM-based judges, which either do not scale or lack reliability for complex, long-horizon manuals. To overcome these limitations, we present MANTRA, a framework for automatically synthesizing machine-checkable compliance benchmarks from natural-language manuals and tool schemas. MANTRA independently generates (i) a symbolic world model capturing procedural dependencies, and (ii) a set of trace-level compliance checks for a given task, and validates their consistency using SMT solving. A structured repair loop resolves inconsistencies, requiring human intervention only as a fallback. %This yields benchmarks that are formally validated. Importantly, MANTRA supports arbitrary domains and long procedural manuals, and provides a tunable notion of task complexity which is utilized to automatically derive challenging tasks accompanying compliance checks. Using MANTRA, we build a new benchmark suite with 285 tasks across 6 domains scaling to 50+ page manuals with minimal human effort. Empirically, we show that the compliance checks are richer with stronger constraint enforcement compared to existing benchmarks. Additionally, the granularity of the checks can be used for debugging the agents' failure modes. These results demonstrate that combining automated benchmark generation with formally grounded validation methods enables scalable and reliable benchmarking of tool-using agents.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
LatentRAG: Latent Reasoning and Retrieval for Efficient Agentic RAG
arXiv:2605.06285v1 Announce Type: new
Abstract: Single-step retrieval-augmented generation (RAG) provides an efficient way to incorporate external information for simple question answering tasks but struggles with complex questions. Agentic RAG extends this paradigm by replacing single-step retrieval with a multi-step process, in which the large language model (LLM) acts as a search agent that generates intermediate thoughts and subqueries to iteratively interact with the retrieval system. This iterative process incurs substantial latency due to the autoregressive generation of lengthy thoughts and subqueries. To address this limitation, we propose LatentRAG, a novel framework that shifts both reasoning and retrieval from discrete language space to continuous latent space. Unlike existing explicit methods that generate natural language thoughts or subqueries token-by-token, LatentRAG produces latent tokens for thoughts and subqueries directly from the hidden states in a single forward pass. We align LLMs with dense retrieval models in the latent space, enabling retrieval over latent subquery tokens and supporting end-to-end joint optimization. To improve transparency and encourage semantically meaningful latent representations, we incorporate a parallel latent decoding mechanism that translates latent tokens back into natural language. Extensive experiments on seven benchmark datasets show that LatentRAG achieves performance comparable to explicit agentic RAG methods while reducing inference latency by approximately 90%, substantially narrowing the latency gap with traditional single-step RAG.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
LEGO: LoRA-Enabled Generator-Oriented Framework for Synthetic Image Detection
arXiv:2605.04445v1 Announce Type: new
Abstract: The rapid advancement of generative technologies has made synthetic images nearly indistinguishable from real ones, thereby creating an urgent need for robust detectors to counter misinformation. However, existing methods mainly rely on universal artifact features that are shared across multiple generators. We observe that as the diversity of generators increases, the overlap of these common features gradually decreases. This severely undermines model generalization. In contrast, focusing only on unique artifacts tends to cause overfitting to specific forgery patterns. To address this challenge, we propose LEGO (LoRA-Enabled Generator-Oriented Framework). The core mechanism of LEGO employs an MLP to modulate multiple LoRA (Low-Rank Adaptation) blocks, each pretrained to capture the unique artifacts of a specific generator, followed by attention-based feature fusion. Unlike conventional methods that seek a single universal solution, LEGO delegates unique artifact extraction to specialized LoRA modules by dividing its training procedure into two stages. Each LoRA module is individually trained on a single-generator dataset to learn generator-specific representations, then MLP and attention layers are trained on mixed datasets to dynamically regulate the contribution of each module. Benefiting from its modular yet robust design, LEGO can be naturally extended by incorporating new LoRA modules for adaptation to newly emerging next-generation datasets, while still achieving substantially better performance than prior SOTA methods with fewer than 30,000 training images, less than 10% of their training data, and only 5 epochs in each training stage.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
Enhancing the interpretability of spatially variable N2O model predictions with soft sensors during wastewater treatment
arXiv:2605.04082v1 Announce Type: new
Abstract: Model-based solutions for nitrous oxide (N2O) emissions from wastewater treatment plants (WWTP) are informed by operational datasets designed to control nutrient levels in liquid waste, coupled with dedicated campaigns for N2O measurements. We analysed how machine learning (ML) models predict disturbances to WWT operation and spatially variable N2O emissions. A real dataset was investigated to validate the modelling framework from N2O emissions predicted by four ML models (R2 = 0.79 - 0.89). Monitoring campaigns for N2O were simulated with a plant-wide mechanistic model to include additional sensors, site-level N2O datasets, and wastewater disturbances (n = 16). ML models were highly accurate (0.97 +- 0.02, n = 80), but the feature importance depended on the model, the scenario and the N2O measurement scale (reactor vs. WWTP). We argue that N2O soft sensor model predictions are limited to the measuring location and the methodological uncertainty of the dataset, which affect the interpretability of the model. Lastly, the analysis of the mechanistic model structure exposed interactions between autotrophic and heterotrophic pathways over nitric oxide which can overestimate aerobic nitrite production and bias the N2O pathway contributions.
Fonte: arXiv cs.LG
RL • Score 85
SOAR: Real-Time Joint Optimization of Order Allocation and Robot Scheduling in Robotic Mobile Fulfillment Systems
arXiv:2605.03842v1 Announce Type: new
Abstract: Robotic Mobile Fulfillment Systems (RMFS) rely on mobile robots for automated inventory transportation, coordinating order allocation and robot scheduling to enhance warehousing efficiency. However, optimizing RMFS is challenging due to strict real-time constraints and the strong coupling of multi-phase decisions. Existing methods either decompose the problem into isolated sub-tasks to guarantee responsiveness at the cost of global optimality, or rely on computationally expensive global optimization models that are unsuitable for dynamic industrial environments. To bridge this gap, we propose SOAR, a unified Deep Reinforcement Learning framework for real-time joint optimization. SOAR transforms order allocation and robot scheduling into a unified process by utilizing soft order allocations as observations. We formulate this as an Event-Driven Markov Decision Process, enabling the agent to perform simultaneous scheduling in response to asynchronous system events. Technically, we employ a Heterogeneous Graph Transformer to encode the warehouse state and integrate phased domain knowledge. Additionally, we incorporate a reward shaping strategy to address sparse feedback in long-horizon tasks. Extensive experiments on synthetic and real-world industrial datasets, in collaboration with Geekplus, demonstrate that SOAR reduces global makespan by 7.5\% and average order completion time by 15.4\% with sub-100ms latency. Furthermore, sim-to-real deployment confirms its practical viability and significant performance gains in production environments. The code is available at https://github.com/200815147/SOAR.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
QKVShare: Quantized KV-Cache Handoff for Multi-Agent On-Device LLMs
arXiv:2605.03884v1 Announce Type: new
Abstract: Multi-agent LLM systems on edge devices need to hand off latent context efficiently, but the practical choices today are expensive re-prefill or full-precision KV transfer. We study QKVShare, a framework for quantized KV-cache handoff between agents that combines token-level mixed-precision allocation, a self-contained CacheCard representation, and a HuggingFace-compatible cache injection path. Our current results support a narrower but clearer story than the original draft: on 150 GSM8K problems with Llama-3.1-8B-Instruct, adaptive quantization remains competitive under repeated handoff and shows its clearest gains against uniform quantization in deeper-hop, higher budget settings; for handoff latency, the QKVShare path reduces TTFT relative to full re prefill at every tested context, from 130.7 ms vs. 150.2 ms at nominal 1K context to 397.1 ms vs. 1029.7 ms at nominal 8K context;. Stage timing shows that post-injection generation, not card creation, dominates the current QKVShare latency path. These results position quantized KV handoff as a promising on-device systems direction while also highlighting the need for stronger controller ablations and apples-to-apples runtime comparisons.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
Learning Time-Varying Graphs from Incomplete Graph Signals
arXiv:2510.17903v3 Announce Type: replace
Abstract: This paper tackles the challenging problem of jointly inferring time-varying network topologies and imputing missing data from partially observed graph signals. We propose a unified non-convex optimization framework to simultaneously recover a sequence of graph Laplacian matrices while reconstructing the unobserved signal entries. Unlike conventional decoupled methods, our integrated approach facilitates a bidirectional flow of information between the graph and signal domains, yielding superior robustness, particularly in high missing-data regimes. To capture realistic network dynamics, we introduce a fused-lasso type regularizer on the sequence of Laplacians. This penalty promotes temporal smoothness by penalizing large successive changes, thereby preventing spurious variations induced by noise while still permitting gradual topological evolution. For solving the joint optimization problem, we develop an efficient Proximal Alternating Direction Method of Multipliers (PADMM) algorithm, which leverages the problem's structure to yield closed-form solutions for both the graph and signal subproblems. This design ensures scalability to large-scale networks and long time horizons. On the theoretical front, despite the inherent non-convexity, we establish a convergence guarantee, proving that the proposed PADMM scheme converges to a stationary point. Furthermore, we derive non-asymptotic statistical guarantees, providing high-probability error bounds for the graph estimator as a function of sample size, signal smoothness, and the intrinsic temporal variability of the graph. Extensive numerical experiments validate the approach, demonstrating that it significantly outperforms state-of-the-art baselines in both convergence speed and the joint accuracy of graph learning and signal recovery.
Fonte: arXiv stat.ML
MLOps/Systems • Score 85
LAWS: Learning from Actual Workloads Symbolically -- A Self-Certifying Parametrized Cache Architecture for Neural Inference, Robotics, and Edge Deployment
arXiv:2605.04069v1 Announce Type: new
Abstract: We introduce LAWS (Learning from Actual Workloads Symbolically), a self-certifying inference caching architecture that builds a growing library of certified expert functions from deployment observations. Each expert covers a region of input space defined by a node in the Probabilistic Language Trie (PLT) of the base model and carries a formal error bound holding uniformly over all inputs. The central result is a self-certification theorem: for any input x, the LAWS approximation error is bounded by epsilon_fit + 2*Lambda(W)*C_E, where Lambda(W) is the model Lipschitz constant, C_E is the maximum embedding diameter, and epsilon_fit is the expert training error -- all checkable at deployment time without ground truth. We prove that LAWS generalizes both Mixture-of-Experts and KV prefix caching as special cases and is strictly more expressive than any fixed-K MoE or finite cache. Further results include a monotone hit rate theorem (any-match routing ensures coverage only increases), an expert library growth rate of O(2^H log N) where H is workload entropy, a fleet learning convergence theorem with Omega(K) speedup for K-unit fleets, and an over-the-air update bandwidth bound. We conjecture that LAWS is acquisition-optimal among stationary online caching algorithms and that the effective Lipschitz constant on the training distribution grows polynomially rather than exponentially in depth. Applications are developed for LLM inference, robotic control, and multi-agent edge deployment.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
AdapShot: Adaptive Many-Shot In-Context Learning with Semantic-Aware KV Cache Reuse
arXiv:2605.03644v1 Announce Type: new
Abstract: Many-Shot In-Context Learning (ICL) has emerged as a promising paradigm, leveraging extensive examples to unlock the reasoning potential of Large Language Models (LLMs). However, existing methods typically rely on a predetermined, fixed number of shots. This static approach often fails to adapt to the varying difficulty of different queries, leading to either insufficient context or interference from noise. Furthermore, the prohibitive computational and memory costs of long contexts severely limit Many-Shot's feasibility. To address the above limitations, we propose AdapShot, which dynamically optimizes shot counts and leverages KV cache reuse for efficient inference. Specifically, we design a probe-based evaluation mechanism that utilizes output entropy to determine the optimal number of shots. To bypass the redundant prefilling computation during both the probing and inference phases, we incorporate a semantics-aware KV cache reuse strategy. Within this reuse strategy, to address positional encoding incompatibilities, we introduce a decoupling and re-encoding method that enables the flexible reordering of cached key-value pairs. Extensive experiments demonstrate that AdapShot achieves an average performance gain of around 10% and a 4.64x speedup compared to state-of-the-art DBSA.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
Multivariate Time Series Data Imputation via Distributionally Robust Regularization
arXiv:2602.00844v2 Announce Type: replace
Abstract: Multivariate time series imputation is often compromised by mismatch between the observed and true data distributions, a bias induced by the combined effects of time-series non-stationarity and systematic missingness. Standard methods that encourage point-wise reconstruction or direct distributional alignment may overfit these biased observations. We propose the Distributionally Robust Regularized Imputer Objective (DRIO), which jointly minimizes reconstruction error and the worst-case divergence between the imputer distribution and data distributions within a Wasserstein ambiguity set. We derive a tractable upper-bound surrogate that reduces infinite-dimensional optimization over measures to adversarial search over sample trajectories, and develop an alternating learning algorithm compatible with modern deep learning backbones. Comprehensive experiments on diverse real-world datasets show that DRIO consistently provides robust imputation and suggests improved downstream forecasting under various missingness scenarios.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Structured Progressive Knowledge Activation for LLM-Driven Neural Architecture Search
arXiv:2605.04057v1 Announce Type: new
Abstract: This paper focuses on a key challenge in Neural Architecture Search (NAS): integrating established architectural knowledge while exploring new designs under expensive evaluations. Large language models (LLMs) are a promising assistant for NAS because they can translate rich architectural and coding priors into executable code edits. However, in practice, seemingly local revisions often propagate into non-local behavioral and performance shifts because a single edit can inadvertently couple multiple interacting functional factors, a phenomenon we refer to as functional entanglement. To make LLM knowledge usable under such entanglement, we propose Structured Progressive Knowledge Activation (SPARK), which activates relevant priors by explicitly selecting the functional factor to modify and conditioning the edit on that factor. This factor-conditioned editing reduces entangled side effects and yields more targeted, reliable architecture modifications. On CLRS-DFS, SPARK achieves a 28.1x sample-efficient architecture evolution speedup and yields a 22.9 percent relative improvement in OOD accuracy.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Storage Is Not Memory: A Retrieval-Centered Architecture for Agent Recall
arXiv:2605.04897v1 Announce Type: new
Abstract: Extraction at ingestion is the wrong primitive for agent memory: content discarded before the query is known cannot be recovered at retrieval time. We propose True Memory, a six-layer architecture that shifts the center of the system from a storage schema to a multi-stage retrieval pipeline operating over events preserved verbatim. The full system runs as a single SQLite file on commodity CPU with no external database, vector index, graph store, or GPU. On LoCoMo (1,540 questions across 10 multi-session conversations), True Memory Pro reaches 93.0% accuracy (3-run mean) against 61.4% for Mem0, 65.4% for Supermemory, approximately 71% for Zep, and 94.5% for EverMemOS under a matched gpt-4.1-mini answer model. On LongMemEval (500 questions), True Memory Pro reaches 87.8% (3-run mean). On BEAM-1M (700 questions at the 1-million-token scale), True Memory Pro reaches 76.6% (3-run mean), above the prior published result of 73.9% for Hindsight. A 56-configuration ablation shows a 1.3-percentage-point spread within the top-performing configuration family.
Fonte: arXiv cs.CL
Theory/Optimization • Score 85
Layerwise LQR for Geometry-Aware Optimization of Deep Networks
arXiv:2605.04230v1 Announce Type: new
Abstract: Geometry-aware optimizers such as Newton and natural gradient can improve conditioning in deep learning, but scalable variants such as K-FAC, Shampoo, and related preconditioners usually impose structural approximations early, often discarding cross-layer interactions induced by the network computation. We introduce Layerwise LQR (LLQR), a framework for learning structured inverse preconditioners under a global layerwise optimal-control objective. The starting point is an exact equivalence: the steepest-descent step under a broad class of divergence-induced quadratic models--including Newton, Gauss-Newton, Fisher/natural-gradient, and intermediate-layer metrics--can be written as a finite-horizon Linear Quadratic Regulator (LQR) problem. This formulation serves as a reference that exposes the layerwise dynamics and cost matrices encoding the original dense geometry. We then derive a scalable relaxation that learns diagonal, (E-)Kronecker-factored, or other structured inverse preconditioners by minimizing the LQR objective and reusing them across iterations. The resulting optimizer wraps standard methods while retaining a principled connection to second-order geometry, without forming or inverting the global curvature matrix. Experiments on ResNets and Transformers show that LLQR improves optimization dynamics and often translates these gains into improved final test performance, while adding only modest wall-clock overhead. It establishes LLQR as a practical framework for geometry-aware second-order methods and a reference for evaluating scalable approximations.
Fonte: arXiv cs.LG
RL • Score 85
Designing a double deep reinforcement learning selection tool for resilient demand prediction
arXiv:2605.04068v1 Announce Type: new
Abstract: The use of artificial intelligence in supply chain forecasting has attracted many scientific studies for several decades. However, the process of selecting an appropriate forecasting solution becomes a daunting task. This complexity arises due to the distinct features inherent to each dataset. Research to tackle this issue has been performed since the eighties but recent development of demand forecasting has opened new perspectives. This research aims to enhance automatic forecasting model selection by proposing a novel architecture that acts as a double deep reinforcement learning agent, selecting automatically a forecasting model from the forecasting committee at the time of prediction. Moreover, a novel early-stopping approach based on average reward convergence has been introduced to expedite training time. To evaluate the model's performance, an empirical study was conducted utilizing grocery sales datasets and snack demands datasets. The experimental results demonstrate the robustness of the proposed approach when compared to state-of-the-art methods.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Stream-T1: Test-Time Scaling for Streaming Video Generation
arXiv:2605.04461v1 Announce Type: new
Abstract: While Test-Time Scaling (TTS) offers a promising direction to enhance video generation without the surging costs of training, current test-time video generation methods based on diffusion models suffer from exorbitant candidate exploration costs and lack temporal guidance. To address these structural bottlenecks, we propose shifting the focus to streaming video generation. We identify that its chunk-level synthesis and few denoising steps are intrinsically suited for TTS, significantly lowering computational overhead while enabling fine-grained temporal control. Driven by this insight, we introduced Stream-T1, a pioneering comprehensive TTS framework exclusively tailored for streaming video generation. Specifically, Stream-T1 is composed of three units: (1) Stream -Scaled Noise Propagation, which actively refines the initial latent noise of the generating chunk using historically proven, high-quality previous chunk noise, effectively establishes temporal dependency and utilizing the historical Gaussian prior to guide the current generation; (2) Stream -Scaled Reward Pruning, which comprehensively evaluates generated candidates to strike an optimal balance between local spatial aesthetics and global temporal coherence by integrating immediate short-term assessments with sliding-window-based long-term evaluations; (3) Stream-Scaled Memory Sinking, which dynamically routes the context evicted from KV-cache into distinct updating pathways guided by the reward feedback, ensuring that previously generated visual information effectively anchors and guides the subsequent video stream. Evaluated on both 5s and 30s comprehensive video benchmarks, Stream-T1 demonstrates profound superiority, significantly improving temporal consistency, motion smoothness, and frame-level visual quality.
Fonte: arXiv cs.CV
Vision • Score 85
UniPCB: A Generation-Assisted Detection Framework for PCB Defect Inspection
arXiv:2605.04635v1 Announce Type: new
Abstract: Printed Circuit Board (PCB) defect inspection faces two compounding challenges: scarce and imbalanced defect samples that limit model training, and insufficient feature representation under complex circuit backgrounds. Existing generation methods rely on single-modality conditions with coarse structural control, while detection methods improve architectures without addressing the data bottleneck. To resolve both challenges jointly, we propose a generation-assisted PCB defect inspection framework that integrates controlled defect synthesis with task-specific defect detection. On the generation side, a Multi-modal Condition Generator extracts complementary edge, depth, and text conditions in parallel. A ScaleEncoder then embeds these conditions into the diffusion U-Net at four resolutions, and a Condition Modulation applies FiLM-style spatially-adaptive modulation at each scale, enabling structurally aligned and defect-aware sample synthesis. On the detection side, an Inverted Residual Shift Attention couples self-attention with shift-wise convolution to jointly capture global context and local texture, and a Cross-level Complementary Fusion Block generates pixel-level gates for selective cross-level feature fusion. The synthesized samples directly enrich the detection training set, so that improvements in generation compound with improvements in detection. Extensive experiments on DsPCBSD+ demonstrate that UniPCB achieves mAP@0.5 of 98.0% and mAP@0.5:0.95 of 61.8% on defect detection, surpassing all compared methods, while the generation branch attains an FID of 129.61 and SSIM of 0.619, outperforming existing conditional generation approaches.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Say the Mission, Execute the Swarm: Agent-Enhanced LLM Reasoning in the Web-of-Drones
arXiv:2605.03788v1 Announce Type: new
Abstract: Large Language Models (LLMs) are increasingly explored as high-level reasoning engines for cyber-physical systems, yet their application to real-time UAV swarm management remains challenging due to heterogeneous interfaces, limited grounding, and the need for long-running closed-loop execution. This paper presents a mission-agnostic, agent-enhanced LLM framework for UAV swarm control, where users express mission objectives in natural language and the system autonomously executes them through grounded, real-time interactions. The proposed architecture combines an LLM-based Agent Core with a Model Context Protocol (MCP) gateway and a Web-of-Drones abstraction based on W3C Web of Things (WoT) standards. By exposing drones, sensors, and services as standardized WoT Things, the framework enables structured tool-based interaction, continuous state observation, and safe actuation without relying on code generation. We evaluate the framework using ArduPilot-based simulation across four swarm missions and six state-of-the-art LLMs. Results show that, despite strong reasoning abilities, current general-purpose LLMs still struggle to achieve reliable execution - even for simple swarm tasks - when operating without explicit grounding and execution support. Task-specific planning tools and runtime guardrails substantially improve robustness, while token consumption alone is not indicative of execution quality or reliability.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
ScrapMem: A Bio-inspired Framework for On-device Personalized Agent Memory via Optical Forgetting
arXiv:2605.03804v1 Announce Type: new
Abstract: Long-term personalized memory for LLM agents is challenging on resource-limited edge devices due to high storage costs and multimodal complexity. To address this, we propose ScrapMem, a framework that integrates multimodal data into "Scrapbook Page." ScrapMem introduces Optical Forgetting, an optical compression mechanism that progressively reduces the resolution of older memories, lowering storage cost while suppressing low-value details. To maintain semantic consistency, we construct an Episodic Memory Graph (EM-Graph) that organizes key events into a causal-temporal structure. Extensive experiments on the multimodal ATM-Bench showcase that ScrapMem provides three main benefits: (1) strong performance, achieving a new state-of-the-art with a 51.0% Joint@10 score; (2) high storage efficiency, reducing memory usage by up to 93% via optical forgetting; and (3) improved recall, increasing Recall@10 to 70.3% through structured aggregation. ScrapMem offers an effective and storage-efficient solution for on-device long-term memory in multimodal LLM agents.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
A Self-Attentive Meta-Optimizer with Group-Adaptive Learning Rates and Weight Decay
arXiv:2605.04055v1 Announce Type: new
Abstract: Adaptive optimizers like AdamW apply uniform hyperparameters across all parameter groups, ignoring heterogeneous optimization dynamics across layers and modules. We address this limitation by proposing MetaAdamW - a new optimizer that integrates a self-attention mechanism to dynamically modulate per-group learning rates and weight decay. The modulation factors are produced by a lightweight Transformer encoder that operates on statistical features (gradient norms, momentum norms, correlations) extracted from each parameter group. To train the attention module, we introduce a meta-learning objective that combines gradient alignment, loss decrease, and generalization gap. A key novel contribution is the extension of homoscedastic uncertainty weighting (HUW) with task-specific priorities that directly scale the regularization terms - enabling domain knowledge to guide automatic loss balancing. Extensive experiments on five diverse tasks-time series forecasting (ETT), language modeling (WikiText-2), machine translation (Multi30k), image classification (CIFAR-10), and sentiment analysis (IMDB) - demonstrate that MetaAdamW consistently outperforms the standard AdamW baseline in terms of validation loss, accuracy, or perplexity. Depending on the task, MetaAdamW either reduces overall training time (by up to 17.11%) or improves performance (by up to 11.08%) while introducing only moderate overhead; in some cases, it can also mitigate issues of insufficient convergence caused by premature early stopping. Ablation studies validate the effectiveness of each component, including feature versions, grouping strategies, and the proposed priority-injected uncertainty weighting.
Fonte: arXiv cs.LG
NLP/LLMs • Score 92
FASQ: Flexible Accelerated Subspace Quantization for Calibration-Free LLM Compression
arXiv:2605.04084v1 Announce Type: new
Abstract: Compressing large language models (LLMs) for deployment on commodity GPUs remains challenging: conventional scalar quantization is limited to fixed bit-widths (e.g., 8/4/3-bit), offers only a few discrete compression points, and typically requires calibration data. We present FASQ (Flexible Accelerated Subspace Quantization), a calibration-free framework that applies product quantization to LLM weight matrices. By tuning two parameters, sub-vector size and codebook cardinality, FASQ exposes a continuous design space spanning 27-49% of the original FP16 model size, filling compression gaps that fixed-bit schemes cannot reach. On Meta-Llama-3-8B, FASQ surpasses 4-bit GPTQ and AWQ in accuracy (67.1-67.7 avg.) at 37-42% model size, with consistent results on Qwen3-8B and Qwen3.5-9B-Base. To make product quantization practical at inference time, we design custom CUDA kernels: a LUT-free direct-compute GEMV for decode and an output-stationary double-buffered LUT GEMM for prefill, both with split-K parallelism. On an RTX~3090, FASQ achieves 45.2 tok/s decode at effective 4-bit (2.56x memory reduction) and 51.8 tok/s at effective 3-bit (2.80x), both surpassing FP16 tensor-core performance (43.9 tok/s) and delivering 1.6 to 1.8x the throughput of AWQ, 2.5 to 2.5x of GPTQ, and 4.3 to 5x of RTN. FASQ is the only compressed method that accelerates decode beyond FP16, offering calibration-free compression, continuous size-quality trade-offs, and real-time inference on a single consumer GPU.
Fonte: arXiv cs.LG
Applications • Score 85
cotomi Act: Learning to Automate Work by Watching You
arXiv:2605.03231v1 Announce Type: new
Abstract: What if a browser agent could learn your work simply by watching you do it? We present cotomi Act, a browser-based computer-using agent that combines reliable multi-step task execution with persistent organizational knowledge learned from user behavior. For execution, an agent scaffold with adaptive lazy observation, verbal-diff-based history compression, coarse-grained actions, and test-time scaling via best-of-N action selection achieves 80.4% on the 179-task WebArena human-evaluation subset, exceeding the reported 78.2% human baseline. For organizational knowledge, a behavior-to-knowledge pipeline passively observes the user's browsing and progressively abstracts it into artifacts (task boards, wiki) exposed through a shared workspace editable by both user and agent. A controlled proxy evaluation confirms that task success improves as behavior-derived knowledge accumulates. In our live demonstration, attendees interact with the system in a real browser, issuing tasks and observing end-to-end autonomous execution and shared knowledge management.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
From Intent to Execution: Composing Agentic Workflows with Agent Recommendation
arXiv:2605.03986v1 Announce Type: new
Abstract: Multi-Agent Systems (MAS) built using AI agents fulfill a variety of user intents that may be used to design and build a family of related applications. However, the creation of such MAS currently involves manual composition of the plan, manual selection of appropriate agents, and manual creation of execution graphs. This paper introduces a framework for the automated creation of multi-agent systems which replaces multiple manual steps with an automated framework. The proposed framework consists of software modules and a workflow to orchestrate the requisite task- specific application. The modules include: an LLM-derived planner, a set of tasks described in natural language, a dynamic call graph, an orchestrator for map agents to tasks, and an agent recommender that finds the most suitable agent(s) from local and global agent registries. The agent recommender uses a two-stage information retrieval (IR) system comprising a fast retriever and an LLM-based re-ranker. We implemented a series of experiments exploring the choice of embedders, re- rankers, agent description enrichment, and supervising critique agent. We benchmarked this system end-to-end, evaluating the combination of planning, agent selection, and task completion, with our proposed approach. Our experimental results show that our approach outperforms the state-of-the- art in terms of the recall rate and is more robust and scalable compared to previous approaches. The critique agent holistically reevaluates both agent and tool recommendations against the overall plan. We show that the inclusion of the critique agent further enhances the recall score, proving that the comprehensive review and revision of task-based agent selection is an essential step in building end-to-end multi-agent systems.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
A Physics-Aware Framework for Short-Term GPU Power Forecasting of AI Data Centers
arXiv:2605.04074v1 Announce Type: new
Abstract: AI data centers experience rapid fluctuations in power demand due to the heterogeneity of computational tasks that they have to support. For example, the power profile of inference and training of large language models (LLMs) is quite distinct and big divergences can result in the instability of the underlying electricity grid. In this paper we propose, to the best of our knowledge, the first physics-informed DLinear time-series model that can accurately forecast power utilization of an AI data center 5-80 minutes (short-term forecasting) into the future. The physics, based on a multi-node lumped thermal resistance-capacitance (RC) network consistent with Newton's law of cooling, is captured using newly derived time-dependent ordinary differential equations (ODE) that separately models and interlinks power consumption with the GPU compute and memory utilization and temperature. The resulting model, that we refer to as PI-DLinear, trained and evaluated on a real AI data center dataset and is not only more accurate than the state-of-the-art (SOTA) models tested, but the forecast profile respects the underlying physics under power throttling and load transient events. Relative to the SOTA transformer-based and non-transformer-based models, improvements in forecasting accuracy (averaged across all look-back and prediction windows) range from 0.782%-39.08% for MSE, 0.993%-51.82% for MAE, and 0.370%-22.28% for RMSE.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
HERCULES: Hardware-Efficient, Robust, Continual Learning Neural Architecture Search
arXiv:2605.04103v1 Announce Type: new
Abstract: Neural Architecture Search (NAS) has emerged as a powerful framework for automatically discovering neural architectures that balance accuracy and efficiency. However, as AI transitions from static benchmarks to real-world deployment, the traditional focus on hardware-aware efficiency is no longer sufficient. We observe that modern NAS methods, especially those that target edge AI, are evolving to address a triple objective: Efficiency, Robustness, and Continual Learning. While efficiency ensures feasibility in resource-constrained environments, robustness guarantees reliability under environmental variabilities, and continual learning enables adaptation to sequential tasks without catastrophic forgetting. We propose a taxonomy of NAS approaches through this triple lens, distinguishing between methods targeting resource optimization, environmental resilience, and architectural plasticity. This unified perspective reveals that these axes, though often studied in isolation, are mutually reinforcing. Building on this taxonomy, we map the current landscape of these NAS methods into a new framework called Hardware-Efficient, Robust, and ContinUal LEarning Search (HERCULES). We define the desiderata, the twelve labours of HERCULES, addressing the non-trivial challenge of balancing an adequate search-space exploration with the immense computational costs of a multi-objective NAS, accounting for these crucial objectives of current AI systems. By identifying critical gaps in existing research, this survey outlines a roadmap toward integrated algorithmic, architectural, and hardware-software co-design for truly deployable, lifelong-learning AI systems.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Replacing Parameters with Preferences: Federated Alignment of Heterogeneous Vision-Language Models
arXiv:2605.03426v1 Announce Type: new
Abstract: Vision-Language Models (VLMs) have broad potential in privacy-sensitive domains such as healthcare and finance, yet strict data-sharing constraints render centralized training infeasible. Federated Learning mitigates this issue by enabling decentralized training, but practical deployments face challenges due to client heterogeneity in computational resources, application requirements, and model architectures. Under extreme model and data heterogeneity, replacing parameter aggregation with preference-based collaboration offers a more suitable interface, as it eliminates the need for direct parameter or data exchange. Motivated by this, we propose MoR, a federated alignment framework that combines GRPO with Mixture-of-Rewards for heterogeneous VLMs. In MoR, each client locally trains a reward model from local preference annotations, capturing specific evaluation signals without exposing raw data. To combine these heterogeneous supervision signals, MoR introduces a Mixture-of-Rewards mechanism with learned routing, which adaptively fuses client reward models according to the input and alignment objective. The server then optimizes a base VLM using GRPO with a KL penalty to a reference model, enabling preference alignment without requiring client models to share architectures or parameters. Experiments on diverse public vision-language benchmarks demonstrate that MoR consistently outperforms federated alignment baselines in generalization and cross-client adaptability. Our approach provides a scalable solution for privacy-preserving alignment of heterogeneous VLMs under federated settings.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
Mechanical Conscience: A Mathematical Framework for Dependability of Machine Intelligenc
arXiv:2605.03847v1 Announce Type: new
Abstract: Distributed collaborative intelligence (DCI), encompassing edge-to-edge architectures, federated learning, transfer learning, and swarm systems, creates environments in which emergent risk is structurally unavoidable: locally correct decisions by individual agents compose into globally unacceptable behavioral trajectories under uncertainty. Existing approaches such as constrained optimization, safe reinforcement learning, and runtime assurance evaluate acceptability at the level of individual actions rather than across behavioral trajectories, and none addresses the multi-participant, uncertainty-laden nature of DCI deployments. This paper introduces mechanical conscience (MC), a novel concept and simplified mathematical framework that operationalizes trajectory-level normative regulation for both single-agent and distributed intelligent systems. Mechanical conscience is defined as a supervisory filter that minimally corrects a baseline policy's actions to reduce cumulative deviation from a normatively admissible region, while accounting for epistemic uncertainty. We introduce associated constructs, conscience score, mechanical guilt, and resonant dependability, that provide an interpretable vocabulary and computable governance signals for this emerging field. Core theoretical properties are established: admissibility equivalence, existence of optimal regulation, and monotonic deviation reduction. Illustrative results demonstrate that MC-regulated agents maintain trajectory-level normative acceptability where conventional controllers drift outside admissible bounds, and that the framework naturally extends to suppress interaction-induced emergent risk in multi-agent DCI settings.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
MP-ISMoE: Mixed-Precision Interactive Side Mixture-of-Experts for Efficient Transfer Learning
arXiv:2605.04058v1 Announce Type: new
Abstract: Parameter-efficient transfer learning (PETL) has emerged as a pivotal paradigm for adapting pre-trained foundation models to downstream tasks, significantly reducing trainable parameters yet suffering from substantial memory overhead caused by gradient backpropagation during fine-tuning. While memory-efficient transfer learning (METL) circumvents this challenge by bypassing backbone gradient computation via lightweight small side networks, its stringent memory constraint severely limits learning capacity of side networks, thereby significantly compromising performance. To address these limitations, we propose a novel Mixed-Precision Interactive Side Mixture-of-Experts framework (MP-ISMoE). Specifically, we first propose a Gaussian Noise Perturbed Iterative Quantization (GNP-IQ) scheme to quantize weights into lower-bits while effectively decreasing quantization errors. By leveraging memory conserved from GNP-IQ, we subsequently employ Interactive Side Mixture-of-Experts (ISMoE) to scaling up side networks without sacrificing overall memory efficiency. Different from conventional mixture-of-experts, ISMoE learns to select optimal experts by interacting with salient features from frozen backbones, thus suppressing knowledge forgetting and boosting performance. Extensive experiments across diverse vision-language and language-only tasks demonstrate that MP-ISMoE remarkably promotes accuracy compared to state-of-the-art METL approaches, while maintaining comparable parameter and memory efficiency.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Learning Correct Behavior from Examples: Validating Sequential Execution in Autonomous Agents
arXiv:2605.03159v1 Announce Type: new
Abstract: As autonomous agents become increasingly sophisticated, validating their sequential behavior presents a significant challenge. Traditional testing approaches require manual specification, exact sequence matching, or thousands of training examples. We present a novel algorithm that automatically learns correct behavior from just 2-10 passing execution traces and validates new executions against this learned model. Our approach combines dominator analysis from compiler theory with multimodal large language model-powered semantic understanding to identify essential states and handle non-deterministic behavior. The system constructs a generalized ground truth model using Prefix Tree Acceptors, merges traces through multi-tiered equivalence detection, and validates new executions via topological subsequence matching. In controlled experiments, our system achieved high accuracy in detecting product bugs and false successes using only 3 training traces. This approach provides explainable validation results with coverage metrics and works across diverse domains including UI testing, code generation, and robotic processes.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
EdgeRazor: A Lightweight Framework for Large Language Models via Mixed-Precision Quantization-Aware Distillation
arXiv:2605.04062v1 Announce Type: new
Abstract: Recent years have witnessed an increasing interest in deploying LLMs on resource-constrained devices, among which quantization has emerged as a promising lightweight technique that converts full-precision model weights and activations into lower-bit formats. Existing weight quantization approaches can be roughly divided into three categories: Post-Training Quantization (PTQ) that calibrates quantized parameters on a small dataset without retraining but suffers from severe performance degradation below 4-bit, Quantization-Aware Training (QAT) that searches low-bit parameters using surrogate gradients but demands substantial computational resources, and Quantization-Aware Distillation that integrates QAT with knowledge transfer from a full-precision teacher but manually selects features to distill and relies heavily on teacher-specific data. In this paper, we propose EdgeRazor, a lightweight framework for LLMs with mixed-precision and extremely low-bit weight quantization. The EdgeRazor framework contains three modules: Mixed-Precision Quantization-Aware Distillation for the fine-grained control of precision, Adaptive Feature Distillation that derives an $n$-bit student from its 16-bit teacher, and Entropy-Aware KL Divergence on both human-annotated and distilled datasets, whose forward-reverse balance is determined solely by the teacher's output distribution. Empirical investigations of EdgeRazor are conducted on base, instruction-tuned, and multimodal LLMs. Notably, EdgeRazor with 1.88-bit surpasses all contenders with the 3-bit precision, especially outperforms the leading 2-bit PTQ methods by 11.3 points, within a 4-10$\times$ lower training budget than the leading QAT approach. EdgeRazor delivers higher compression ratios at all bit width; the 1.58-bit Qwen3-0.6B reduces storage from 1.41 GB to 0.28 GB while accelerating decoding by 15.1$\times$ relative to the 16-bit baseline.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Lightning Unified Video Editing via In-Context Sparse Attention
arXiv:2605.04569v1 Announce Type: new
Abstract: Video editing has evolved toward In-Context Learning (ICL) paradigms, yet the resulting quadratic attention costs create a critical computational bottleneck. In this work, we propose In-context Sparse Attention (ISA), the first near-lossless empirical sparse framework tailored for ICL video editing. Our design is grounded in two key insights: first, context tokens exhibit significantly lower saliency than source tokens; second, we theoretically prove and empirically validate that Query sharpness correlates with approximation error. Motivated by these findings, ISA implements an efficient pre-selection strategy to prune redundant context, followed by a dynamic query grouping mechanism that routes high-error queries to full attention and low-error ones to a computationally efficient 0-th order Taylor sparse attention. Furthermore, we build \textbf{\texttt{LIVEditor}} , a novel lightning video editing model via ISA and a proposed video-editing data pipeline that curated a 1.7M high-quality dataset. Extensive experiments demonstrate that LIVEditor achieves a $\sim$60% reduction in attention-module latency while surpassing state-of-the-art methods across EditVerseBench, IVE-Bench, and VIE-Bench, delivering near-lossless acceleration without compromising visual fidelity.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Rethinking Local Learning: A Cheaper and Faster Recipe for LLM Post-Training
arXiv:2605.04913v1 Announce Type: new
Abstract: LLM post-training typically propagates task gradients through the full depth of the model. Although this end-to-end structure is simple and general, it couples task adaptation to full-depth activation storage, long-range backward dependencies and direct task-gradient access to pretrained representations. We argue that this full-depth backward coupling can be unnecessarily expensive and intrusive, particularly when post-training supervision is much narrower than pre-training. To this end, we propose \textbf{LoPT}: Local-Learning Post-Training, a simple post-training strategy that makes gradient reach an explicit design choice. LoPT places a single gradient boundary at the transformer midpoint: the second-half block learns from the task objective, while the first-half block is updated by a lightweight feature-reconstruction objective to preserve useful representations and maintain interface compatibility. LoPT shortens the task-induced backward path while limiting direct interference from narrow task gradients on early-layer representations. Extensive experiments demonstrate that LoPT achieves competitive performance with lower memory cost, higher training efficiency and better retention of pretrained capabilities. Our code is available at: https://github.com/HumyuShi/LoPT
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
Temporal Structure Matters for Efficient Test-Time Adaptation in Wearable Human Activity Recognition
arXiv:2605.04617v1 Announce Type: new
Abstract: Wearable human activity recognition (WHAR) models often suffer from performance degradation under real-world cross-user distribution shifts. Test-time adaptation (TTA) mitigates this degradation by adapting models online using unlabeled test streams, yet existing methods largely inherit assumptions from vision tasks and underexploit the inherent inter-window temporal structure in WHAR streams. In this paper, we revisit such temporal structure as a feature-conditioned inference signal rather than merely an output-space smoothing prior. We derive the insight that temporal continuity and observation-induced feature deviations provide complementary cues for determining when to preserve or release temporal inertia and where to route prediction refinement during likely transitions. Building upon this insight, we propose SIGHT, a lightweight and backpropagation-free TTA framework for WHAR, enabling real-time edge deployment. SIGHT estimates predictive surprise by comparing the current feature with a prototype-based expected state, and then uses the resulting feature deviation to guide geometry-aware transition routing based on prototype alignment and stream-level marginal habit tracking. Evaluations on real-world datasets confirm that SIGHT outperforms existing TTA baselines while reducing computational and memory costs.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
HUGO-CS: A Hybrid-Labeled, Uncertainty-Aware, General-Purpose, Observational Dataset for Cold Spray
arXiv:2605.04257v1 Announce Type: new
Abstract: Cold spraying is an increasingly common approach for repairing and manufacturing components due to its solid-state manufacturing capabilities. However, process optimization remains difficult due to many interdependent parameters and the lack of large-scale, machine-readable data to support modeling. While the scientific literature contains many relevant experiments, results are inconsistently reported (often in tables and figures) and use non-uniform units, limiting utilization at scale. To address these limitations, this work presents HUGO-CS, a literature-derived dataset of 4,383 cold-spray experiments with 144 features from 1,124 sources, exceeding the previous largest dataset (137 samples) by 30x. With completely manual extraction requiring an average of 91 minutes per document, this work designs and leverages a Hybrid-labeled, Uncertainty-aware, General-purpose, Observational extraction framework, called HUGO, to support this extraction. HUGO combines automated LLM-based labeling with targeted manual label refinement to handle this experimental result extraction process from scientific literature. To balance labeling efficiency with extraction accuracy, HUGO introduces a Hierarchical Risk Mitigation (HRM) to route LLM outputs with a high risk of potential errors for manual review, while retaining low-risk records as auto-labeled. Lastly, HUGO post-processing consolidates categorical descriptors, maps reported feedstock chemistries into structured continuous compositions, and normalizes units across sources. Of the 4,383 reported experiments, 1,765 are hand-labeled, providing a high-quality labeled subset for benchmarking, error analysis, and higher-fidelity data points. All code to replicate this work, along with the complete HUGO-CS dataset, are released under a CC-BY license at https://github.com/sprice134/HUGO.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
FL-Sailer: Efficient and Privacy-Preserving Federated Learning for Scalable Single-Cell Epigenetic Data Analysis via Adaptive Sampling
arXiv:2605.04519v1 Announce Type: cross
Abstract: Single-cell ATAC-seq (scATAC-seq) enables high-resolution mapping of chromatin accessibility, yet privacy regulations and data size constraints hinder multi-institutional sharing. Federated learning (FL) offers a privacy-preserving alternative, but faces three fundamental barriers in scATAC-seq analysis: ultra-high dimensionality, extreme sparsity, and severe cross-institutional heterogeneity. We propose FL-Sailer, the first FL framework designed for scATAC-seq data. FL-Sailer integrates two key innovations: (i) adaptive leverage score sampling, which selects biologically interpretable features while reducing dimensionality by 80%, and (ii) an invariant VAE architecture, which disentangles biological signals from technical confounders via mutual information minimization. We provide a convergence guarantee, showing that FL-Sailer converges to an approximate solution of the original high-dimensional problem with bounded error. Extensive experiments on synthetic and real epigenomic datasets demonstrate that FL-Sailer not only enables previously infeasible multi-institutional collaborations but also surpasses centralized methods by leveraging adaptive sampling as an implicit regularizer to suppress technical noise. Our work establishes that federated learning, when tailored to domain-specific challenges, can become a superior paradigm for collaborative epigenomic research.
Fonte: arXiv stat.ML
MLOps/Systems • Score 85
Road Risk Monitor: A Deployable U.S. Road Incident Forecasting System with Live Weather and Road-Level Tiles
arXiv:2605.04242v1 Announce Type: new
Abstract: Nationwide road-incident forecasting is a systems problem before it is a modeling problem. A usable service must connect historical incident archives, historicalandliveweather,nationalroadgeometry, offline model training, tile generation, web serving and runtime handoff. This paper presents Road Risk Monitor, a U.S.-wide road-safety stack that combines a nationwide H3 baseline trained on FARS fatal-crash data with a road-segment forecasting pipeline trained from TIGER/Line geometry and US-Accidents events, then serves predictions through live APIs, raster tiles, JSON road tiles, and a public web application.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
MEMTIER: Tiered Memory Architecture and Retrieval Bottleneck Analysis for Long-Running Autonomous AI Agents
arXiv:2605.03675v1 Announce Type: new
Abstract: Long-running autonomous AI agents suffer from a well-documented memory coherence problem: tool-execution success rates degrade 14 percentage points over 72-hour operation windows due to four compounding failure modes in existing flat-file memory systems. We present MEMTIER, a tripartite memory architecture for the OpenClaw agent runtime that introduces a structured episodic JSONL store, a five-signal weighted retrieval engine, an attention-attributed cognitive weight update loop, an asynchronous consolidation daemon promoting episodic facts to a semantic tier, and a PPO-based policy framework for adapting retrieval weights (infrastructure validated; performance gains pending camera-ready). On the full 500-question LongMemEval-S benchmark (Wu et al., 2025), MEMTIER achieves Acc=0.382, F1=0.412 with Qwen2.5-7B on a consumer 6GB GPU - a +33 percentage point improvement over the full-context baseline (0.050 -> 0.382, i.e., 5% -> 38%). With DeepSeek-V4-Flash fact pre-population, single-session recall reaches 0.686-0.714, exceeding the paper's RAG BM25 GPT-4o baseline (0.560) on those categories. Temporal reasoning rises to 0.323 and multi-session synthesis to 0.173, demonstrating that structured semantic pre-population qualitatively changes what lightweight retrieval can achieve. All phases run locally on a consumer laptop with a 6GB GPU.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Robust Agent Compensation (RAC): Teaching AI Agents to Compensate
arXiv:2605.03409v1 Announce Type: new
Abstract: We present Robust Agent Compensation (RAC), a log-based recovery paradigm (providing a safety net) implemented through an architectural extension that can be applied to most Agent frameworks to support reliable executions (avoiding unintended side effects). Users can choose to enable RAC without changing their current agent code (e.g., LangGraph agents). The proposed approach can be implemented in most existing agent frameworks via their existing extension points. We present an implementation based on LangChain, demonstrate its viability through the $\tau$-bench and REALM-Bench, and show that when solving complex problems, RAC is 1.5-8X or more better in both latency and token economy compared to state-of-the-art LLM-based recovery approaches.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Predict-then-Diffuse: Adaptive Response Length for Compute-Budgeted Inference in Diffusion LLMs
arXiv:2605.04215v1 Announce Type: new
Abstract: Diffusion-based Large Language Models (D-LLMs) represent a promising frontier in generative AI, offering fully parallel token generation that can lead to significant throughput advantages and superior GPU utilization over traditional autoregressive paradigm. However, this parallelism is constrained by the requirement of a fixed-size response length prior to generation. This architectural limitation imposes a severe trade-off: oversized response length results in computational waste on semantically meaningless padding tokens, while undersized response length cause output truncation requiring costly re-computations that introduce unpredictable latency spikes. To tackle this issue, we propose Predict-then-Diffuse, a simple and model-agnostic framework, that enables compute-budgeted inference per input query by first estimating the response length and then using it to run inference with D-LLM. At its core lies a Adaptive Response Length Predictor (AdaRLP) auxiliary predictor that predicts the optimal response length given an input query. As a measure against under-predicting the response length and re-running inference with a higher response length, we introduce a data-driven safety mechanism, which trades a negligible padding overhead. As a whole, our framework limits the significant waste of computation on padding tokens and preserves output quality. Experimental validation on multiple datasets demonstrate that Predict-then-Diffuse significantly reduces computational costs (FLOP) compared to the default D-LLM inference mechanism and baselines based on heuristics, while being robust to skewed data distributions.
Fonte: arXiv cs.LG
RL • Score 85
Explaining and Preventing Alignment Collapse in Iterative RLHF
arXiv:2605.04266v1 Announce Type: new
Abstract: Reinforcement learning from human feedback (RLHF) typically assumes a static or non-strategic reward model (RM). In iterative deployment, however, the policy generates the data on which the RM is retrained, creating a feedback loop. Building on the Stackelberg game formulation of this interaction, we derive an analytical decomposition of the policy's true optimization gradient into a standard policy gradient and a parameter-steering term that captures the policy's influence on the RM's future parameters. We show that standard iterative RLHF, which drops this steering term entirely, suffers from alignment collapse: the policy systematically exploits the RM's blind spots, producing low-quality, high-reward outputs whose feedback reinforces the very errors it exploits. To mitigate this, we propose foresighted policy optimization (FPO), a mechanism-design intervention that restores the missing steering term by regularizing the policy's parameter-steering effect on RM updates. We instantiate FPO via a scalable first-order approximation and demonstrate that it prevents alignment collapse on both controlled environments and an LLM alignment pipeline using Llama-3.2-1B.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
MuCALD-SplitFed: Causal-Latent Diffusion for Privacy-Preserving Multi-Task Split-Federated Medical Image Segmentation
arXiv:2605.04108v1 Announce Type: new
Abstract: Federated Learning enables decentralized training by aggregating model updates across clients without sharing raw data, while Split Federated Learning further partitions the model between clients and a server to reduce computation and communication at the client side. However, decentralized medical institutions rarely operate on a single shared task, making standard Federated and SplitFed collaborations poorly aligned with real clinical workflows. Multi-task FL extends these frameworks by allowing clients to handle different tasks, but often introduces instability and privacy vulnerabilities. This study proposes \textbf{MuCALD-SplitFed}, a multi-task SplitFed framework that integrates causal representation learning and latent diffusion. Experiments show MuCALD-SplitFed consistently improves segmentation, while baseline SplitFed fails to converge. The proposed approach further reduces information leakage at split points, mitigating reconstruction-based and membership inference attacks. Additionally, MuCALD SplitFed outperforms state-of-the-art personalized FL and multi-task FL approaches. The code repository is: https://github.com/ChamaniS/MuCALD_SplitFed.
Fonte: arXiv cs.CV
Theory/Optimization • Score 85
Estimating the expected output of wide random MLPs more efficiently than sampling
arXiv:2605.05179v1 Announce Type: cross
Abstract: By far the most common way to estimate an expected loss in machine learning is to draw samples, compute the loss on each one, and take the empirical average. However, sampling is not necessarily optimal. Given an MLP at initialization, we show how to estimate its expected output over Gaussian inputs without running samples through the network at all. Instead, we produce approximate representations of the distributions of activations at each layer, leveraging tools such as cumulants and Hermite expansions. We show both theoretically and empirically that for sufficiently wide networks, our estimator achieves a target mean squared error using substantially fewer FLOPs than Monte Carlo sampling. We find moreover that our methods perform particularly well at estimating the probabilities of rare events, and additionally demonstrate how they can be used for model training. Together, these findings suggest a path to producing models with a greatly reduced probability of catastrophic tail risks.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
RemoteZero: Geospatial Reasoning with Zero Human Annotations
arXiv:2605.04451v1 Announce Type: new
Abstract: Geospatial reasoning requires models to resolve complex spatial semantics and user intent into precise target locations for Earth observation. Recent progress has liberated the reasoning path from manual curation, allowing models to generate their own inference chains. Yet a final dependency remains: they are still supervised by human-annotated ground-truth coordinates. This leaves the reasoning process autonomous, but not its spatial endpoint, and prevents true self-evolution on abundant unlabeled remote sensing data. To break this bottleneck, we introduce RemoteZero, a box-supervision-free framework for geospatial reasoning. RemoteZero is motivated by a simple asymmetry: an MLLM is typically better at verifying whether a region satisfies a query than at directly generating precise coordinates. Leveraging this stronger discriminative ability, RemoteZero replaces geometric supervision with intrinsic semantic verification and enables GRPO training without box annotations. The resulting framework further supports iterative self-evolution, allowing the model to improve from unlabeled remote sensing imagery through its own verification signal. Experiments show that RemoteZero achieves competitive performance against strong supervised methods, demonstrating the potential of self-verifying training for geospatial reasoning localization.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
Disentangled Learning Improves Implicit Neural Representations for Medical Reconstruction
arXiv:2605.04234v1 Announce Type: new
Abstract: Implicit neural representations (INRs) have emerged as a powerful paradigm for medical imaging via physics-informed unsupervised learning. Classical INRs optimize an entire network from scratch for each subject, leading to inefficient training and suboptimal imaging quality. Recent initialization-based approaches attempt to inject population priors into pre-trained networks, yet they rely on high-quality images and often suffer from catastrophic forgetting during fine-tuning. We present DisINR, a novel INR framework that explicitly disentangles shared and subject-specific representations. DisINR introduces a shared encoder-decoder pair and subject-specific encoders, whose features are jointly decoded for image reconstruction. By integrating differentiable forward models, it pre-trains the shared modules directly from limited raw measurements, removing the need for pre-acquired high-quality images. During test-time adaptation, only the subject-specific encoder is optimized, while the shared pair remains frozen, effectively preserving learned priors. Extensive evaluations on three representative medical imaging tasks show that DisINR significantly outperforms state-of-the-art INRs in both reconstruction accuracy and efficiency.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
Accelerating battery research with an AI interface between FINALES and Kadi4Mat
arXiv:2605.00909v1 Announce Type: new
Abstract: The time-consuming formation process critically impacts the longevity of sodium-ion coin cells and End Of Life (EOL) performance. This study aims to optimize formation protocols for duration efficiency, targeting high-performance outcomes while minimizing the number of experiments to reduce resource consumption and accelerate discovery. Specifically, we consider two potentially competing objectives: minimizing formation time and maximizing EOL performance. Beyond this application focus, we also present a methodological contribution: a framework designed to enable interoperability between the FINALES and Kadi RDM ecosystems, which we employ to tackle our optimization problem. In this setup, the FINALES framework orchestrates experiment planning and execution on the POLiS MAP, while an active-learning agent implemented within Kadi4Mat guides experiment selection, using multi-objective batched Bayesian optimization to efficiently explore the parameter space. This interoperability enhancement enables coordinated, distributed collaboration across automated systems and human-operated workflows, bridging multiple research centers. Using this approach, we iteratively explore the trade-off between formation time and EOL performance and identify candidate solutions approximating the Pareto front. The resulting workflow demonstrates the capability of interoperable infrastructures to facilitate data-driven optimization in battery research, and establishes a transferable framework applicable to diverse materials science and engineering optimization tasks.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Joint Energy Management and Coordinated AIGC Workload Scheduling for Distributed Data Centers: A Diffusion-Aided Reward Shaping Approach
arXiv:2605.02965v1 Announce Type: new
Abstract: Artificial intelligence-generated content (AIGC) has emerged as a transformative paradigm for automating the creation of diverse and customized content, giving rise to rapidly growing computational workloads in cloud data centers. It is imperative for AIGC service providers (ASPs) to strategically schedule AIGC workloads to reduce data center energy costs while guaranteeing high-quality content generation. However, the distinctive characteristics of AIGC services pose critical challenges, including model heterogeneity across ASPs, implicit service quality evaluation, and complex inference process control. To tackle these challenges, we propose a joint energy management and coordinated AIGC workload scheduling framework, which introduces an explicit mathematical characterization of service quality to promote both job transfer among ASPs and fine-grained inference process configuration. Moreover, various energy resources within data centers are jointly considered to enhance power usage flexibility. Subsequently, a system utility maximization problem is formulated to balance AIGC service revenue with operational penalties and costs. Nevertheless, the strong coupling among job scheduling decisions induces severe reward sparsity, which limits the effectiveness of existing deep reinforcement learning (DRL) algorithms. To address this issue, we develop a diffusion model-aided reward shaping approach to synthesize complementary reward signals through a multi-step denoising process. This approach is seamlessly integrated with DRL to enable efficient learning of scheduling policies under sparse environmental feedback. Experiments based on real-world models and datasets demonstrate that our scheme effectively accommodates electricity price fluctuations and AIGC model heterogeneity, while achieving superior learning convergence and system utility compared with benchmark methods.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Adaptive Data Compression and Reconstruction for Memory-Bounded EEG Continual Learning
arXiv:2605.03085v1 Announce Type: new
Abstract: Electroencephalography (EEG) signals provide millisecond-level temporal resolution but their analysis is limited by remarkable noise and inter-subject variability, making robust personalization difficult under limited annotations. Unsupervised Individual Continual Learning (UICL) has been proposed to address this practical challenge, where a model pretrained on a labeled cohort must adapt online to unlabeled subject streams under strict memory constraints. However, existing UICL methods typically store full past samples, which undermine the continual learning goal of avoiding retraining. Observing that EEG signals exhibit well-structured morphologies to be exploited via morphology-aware selection, compression, and reconstruction, here we propose Adaptive Data Compression and Reconstruction (ADaCoRe) for UICL. This is a memory-efficient pipeline composed of saliency-driven keyframe protection, rational polyphase compression, adjoint reconstruction with verbatim overwrite on protected indices, and prototype-confidence selection for adaptive exemplar maintenance. Across three representative benchmarks, ADaCoRe consistently outperforms recent strong baselines under tight buffer regimes (eg., the performance gains are at least +2.7 and +15.3 ACC on ISRUC and FACED datasets, respectively). Ablation studies quantify compression-fidelity trade-offs and highlight the contribution of each design, while visualizations confirm the preservation of key EEG morphology during compression and reconstruction.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Mixed-Precision Information Bottlenecks for On-Device Trait-State Disentanglement in Bipolar Agitation Detection
arXiv:2605.03039v1 Announce Type: new
Abstract: Continuous monitoring of bipolar disorder agitation via voice biomarkers requires disentangling stable speaker traits from volatile affective states on resource-constrained edge devices. We introduce MP-IB, the first framework to treat mixed-precision quantization as an information bottleneck for clinical trait-state separation. The core insight is that numerical precision itself controls capacity: an FP16 trait head (1,024 bits) encodes speaker identity, while an INT4 state head (128 bits) captures agitation, yielding 8x information asymmetry without adversarial training. We augment this with Dynamic Precision Scheduling and Multi-Scale Temporal Fusion. On Bridge2AI-Voice (N=833, 4 sessions/participant, strict speaker-independent CV), MP-IB achieves rho = 0.117 (95\% CI: [0.089, 0.145], p=0.003 vs. chance), outperforming 94M-parameter WavLM-Adapter with in-domain SSL continuation (rho = -0.042), beta VAE disentanglement (rho = 0.089), and hand-crafted prosody (rho = 0.031) by 2.8--15.9 points absolute. Zero-shot transfer to CREMA-D achieves AUC=0.817. Identity leakage is suppressed to near-random (EER=0.42, MIA-AUC=0.52). End-to-end latency is 23.4 ms with a 617 KB footprint, enabling real-time monitoring on sub 20 dollar devices.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Low Rank Tensor Completion via Adaptive ADMM
arXiv:2605.03736v1 Announce Type: new
Abstract: We consider a novel algorithm, for the completion of partially observed low-rank tensors, as a generalization of matrix completion. The proposed low-rank tensor completion (TC) method builds on the conventional nuclear norm (NN) minimization-based low-rank TC paradigm, by leveraging the alternating direction method of multipliers (ADMM) optimization framework. To that extend the original NN minimization problem is reformulated into multiple subproblems, which are then solved iteratively via closed-form proximal operators, making use of over-relaxation and an adaptive penalty parameter update scheme, to further speed up convergence and improve the overall performance of the method. Simulation results demonstrate the superior performance of the new method in terms of normalized mean square error (NMSE), compared to the conventional state-of-the-art (SotA) techniques, including NN minimization approaches, as well as a mixture of the latter with a matrix factorization approach, while its convergence can be significantly improved by initializing the algorithm with the solution of the SotA.
Fonte: arXiv stat.ML
MLOps/Systems • Score 85
StateSMix: Online Lossless Compression via Mamba State Space Models and Sparse N-gram Context Mixing
arXiv:2605.02904v1 Announce Type: new
Abstract: We present StateSMix, a fully self-contained lossless compressor that couples an online-trained Mamba-style State Space Model (SSM) with sparse n-gram context mixing and arithmetic coding. The model is initialised from scratch and trained token-by-token on the file being compressed, requiring no pre-trained weights, no GPU, and no external dependencies. The SSM (DM=32, NL=2, approximately 120K active parameters per file) provides a continuously-updated probability estimate over BPE tokens, while nine sparse n-gram hash tables (bigram through 32-gram, 16M slots each) add exact local and long-range pattern memorisation via a softmax-invariant logit-bias mechanism that updates only non-zero-count tokens. An entropy-adaptive scaling mechanism modulates the n-gram contribution based on the SSM's predictive confidence, preventing over-correction when the neural model is already well-calibrated. On the standard enwik8 benchmark, StateSMix achieves 2.123 bpb on 1 MB, 2.149 bpb on 3 MB, and 2.162 bpb on 10 MB, beating xz -9e (LZMA2) by 8.7%, 5.4%, and 0.7% respectively. Ablation experiments establish the SSM as the dominant compression engine: it alone accounts for a 46.6% size reduction over a frequency-count baseline and beats xz without any n-gram component, while n-gram tables provide a complementary 4.1% gain through exact context memorisation. OpenMP parallelisation of the training loop yields 1.9x speedup on 4 cores. The system is implemented in pure C with AVX2 SIMD and processes approximately 2,000 tokens per second on commodity x86-64 hardware.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Real Image Denoising with Knowledge Distillation for High-Performance Mobile NPUs
arXiv:2605.03680v1 Announce Type: new
Abstract: While deep-learning-based image restoration has achieved unprecedented fidelity, deployment on mobile Neural Processing Units (NPUs) remains bottlenecked by operator incompatibility and memory-access overhead. We propose an NPU-aware hardware-algorithm co-design approach for real-world image denoising on mobile NPUs. Our approach employs a high-capacity teacher to supervise a lightweight student network specifically designed to leverage the tiled-memory architectures of modern mobile SoCs. By prioritizing NPU-native primitives -- standard 3x3 convolutions, ReLU activations, and nearest-neighbor upsampling -- and employing a progressive context expansion strategy (up to 1024x1024 crops), the model achieves 37.66 dB PSNR / 0.9278 SSIM on the validation benchmark and 37.58 dB PSNR / 0.9098 SSIM on the held-out test benchmark at full resolution (2432x3200) in the Mobile AI 2026 challenge. Following the official challenge rules, the inference runtime is measured under a standardized Full HD (1088x1920) protocol, where it runs in 34.0 ms on the MediaTek Dimensity 9500 and 46.1 ms on the Qualcomm Snapdragon 8 Elite NPU. We further reveal an "Inference Inversion" effect, where strict adherence to NPU-compatible operations enables dedicated NPU execution up to 3.88x faster than the integrated mobile GPU. The 1.96M-parameter student recovers 99.8% of the teacher's restoration quality via high-alpha knowledge distillation (alpha = 0.9), achieving a 21.2x parameter reduction while closing the PSNR gap from 1.63 dB to only 0.05 dB. These results establish hardware-aware distillation as an effective strategy for unifying high-fidelity denoising with practical deployment across diverse mobile NPU architectures. The proposed lightweight student model (LiteDenoiseNet) and its training statistics are provided in the NN Dataset, available at https://github.com/ABrain-One/NN-Dataset.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
ZeRO-Prefill: Zero Redundancy Overheads in MoE Prefill Serving
arXiv:2605.02960v1 Announce Type: new
Abstract: Production LLM workloads increasingly serve discriminative tasks, such as classification, recommendation, and verification, whose answers are read from the logits of a single prefill pass with no autoregressive decoding. Serving these prefill-only workloads on mixture-of-experts (MoE) models is bottlenecked not by compute but by the distributed execution required to fit the model: existing parallel strategies (tensor, expert, and pipeline parallelism) trade memory pressure for redundant computation, communication, and synchronization, severely degrading MoE prefill serving efficiency. We observe that these overheads stem from coupling expert placement with synchronous activation routing -- a design inherited from the decoding era. The long, compute-bound forward passes of large-batch prefill open a per-layer window wide enough to stream expert weights in the background, replacing per-layer activation AllToAll with asynchronous weight AllGather fully overlapped with computation. We propose ZeRO-Prefill, a prefill-only serving system whose backend, AsyncEP (Asynchronous Expert Parallelism), gathers experts by weight rather than routing them by activation, and whose frontend co-enforces a physically-derived saturation threshold through prefix-aware routing and true-FLOPs load tracking. On Qwen3-235B-A22B across four hardware/precision configurations, ZeRO-Prefill delivers 1.35-1.37x throughput over the strongest distributed baseline on real-world workloads and up to 1.59x on long-context synthetic workloads, sustaining 29.8-36.2% per-GPU model FLOPs utilization.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
NH-CROP: Robust Pricing for Governed Language Data Assets under Cost Uncertainty
arXiv:2605.01745v1 Announce Type: new
Abstract: Language data are increasingly acquired and governed as assets, yet platforms often price candidate resources before knowing their true privacy or access costs. We study online pricing for governed language data assets under cost uncertainty. At each round, a platform observes an NLP task, a candidate asset, and a coarse cost estimate, may pay for a refined cost signal, posts a price, and receives safe net revenue.
We introduce \textsc{NH-CROP}, a clipped robust pricing framework with a no-harm information-acquisition gate. The method compares direct pricing, risk-aware pricing, and verify-then-price, and acquires information only when its estimated decision value exceeds the best no-verification alternative. Across synthetic, real-proxy, and downstream-utility-grounded benchmarks, clipped \textsc{NH-CROP} variants improve or remain competitive with price-only and risk-aware baselines. Causal ablations show that paid verification is not the main source of gains in real-proxy and utility-grounded settings: the strongest learned policies often choose not to verify. Oracle and high-decision-value diagnostics show that refined cost information can still have substantial local value. Overall, governed language-data platforms should calibrate pricing under uncertain access costs first and verify only when information is cheap and decision-actionable.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
Agentic AI Systems Should Be Designed as Marginal Token Allocators
arXiv:2605.01214v1 Announce Type: new
Abstract: This position paper argues that agentic AI systems should be designed and evaluated as \emph{marginal token allocation economies} rather than as text generators priced by the unit. We follow a single request -- a developer asking a coding agent to fix a failing test -- through four economic layers that today are designed in isolation: a router that decides which model answers, an agent that decides whether to plan, act, verify, or defer, a serving stack that decides how to produce each token, and a training pipeline that decides whether the trace is worth learning from. We show that all four layers are solving the \emph{same} first-order condition -- marginal benefit equals marginal cost plus latency cost plus risk cost -- with different index sets and different prices. The framing is deliberately minimal: we do not propose a complete theory of AI economics. But adopting marginal token allocation as the shared accounting object explains why systems that locally minimize tokens globally misallocate them, predicts a small set of recurring failure modes (over-routing, over-delegation, under-verification, serving congestion, stale rollouts, cache misuse), and points to a concrete research agenda in token-aware evaluation, autonomy pricing, congestion-priced serving, and risk-adjusted RL budgeting.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Model Routing as a Trust Problem: Route Receipts for Adaptive AI Systems
arXiv:2605.01710v1 Announce Type: new
Abstract: AI products often route requests through version aliases, service tiers, tool choices, regional endpoints, fallback rules, or safety handling before responding. These routing steps are documented product surfaces in several widely used AI platforms and serving stacks.
Routing helps AI services stay affordable, fast, and available at scale, and it shapes trust. Trust can break when routing changes the cost, quality, or accountability of a response without the user being able to tell what happened. "Which model answered?" is only part of the audit question. The runtime path matters.
Adaptive AI systems should produce a runtime transparency artifact called the route receipt. A route receipt is a compact record of the route that served a request. It should capture enough material facts for people relying on the output to reconstruct important routing decisions without exposing proprietary internals or hidden reasoning.
Route transparency should be part of model documentation. Model cards describe trained model artifacts, while route receipts describe the runtime conditions under which a particular answer was produced. The paper introduces the route-receipt concept, a minimal schema and redaction model, and a documentation-based survey of selected platforms showing that receipt fragments already exist without a portable per-answer record.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Cascade Token Selection for Transformer Attention Acceleration
arXiv:2605.03110v1 Announce Type: new
Abstract: A method is presented for reducing the cost of representative token selection in transformer attention layers by exploiting the coherence of the representative set across depth. Activation Decorrelation Attention (ADA) selects $r \ll T$ representative tokens at each layer via a Gram threshold and computes attention on the compressed $r \times r$ problem, but the selection requires a $T \times T$ Gram matrix at every layer. The cascade mechanism introduced here inherits the representative set from layer $l$ to layer $l+1$, validates it via a $(T - r) \times r$ cross-Gram computation, and updates it with a small number of additions and removals. The cost of the selection step drops from $O(T^2 d)$ to $O(T r d)$ per layer. Validation on three model families (GPT-2 124M, GPT-J 6B, OPT 6.7B) on AMD MI300X demonstrates Gram operation savings of $22\%$ to $63\%$ with mean Jaccard overlap of $0.83$ to $0.94$ between consecutive layers. The cascade reveals that the set of informative tokens is a structural property of the input that propagates coherently through the depth of the network: the same tokens carry the non-redundant information at layer $l$ and at layer $l+1$.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Agentic AI-Based Joint Computing and Networking via Mixture of Experts and Large Language Models
arXiv:2605.02911v1 Announce Type: new
Abstract: Future sixth-generation (6G) mobile networks are envisioned to be equipped with a diverse set of powerful, yet highly specialized, optimization experts. Such a promising vision is concurrently expected to give rise to the need for scalable mechanisms that can select, combine, and orchestrate such experts based on high-level intent and uncertainty descriptions. In this paper, we propose an agentic artificial intelligence (AI)-based network optimization framework that integrates mixture of experts (MoE) architectures with large language models (LLMs). Under the proposed framework, the employed LLM acts as a semantic gate to reason over operator objectives and dynamically compose suitable optimization agents. The proposed framework is formulated in a model-agnostic manner and bridges human-readable network intents with low-level resource allocation decisions, enabling flexible optimization across heterogeneous objectives and operating conditions. As a representative instantiation, we apply the framework to a joint communication and computing network and design a library of specialized optimization experts covering throughput, fairness, and delay-driven objectives under both regular and robust conditions. Numerical simulations demonstrate that the proposed agentic MoE framework consistently achieves near-optimal performance compared to exhaustive expert combinations while outperforming individual experts across diverse objectives, including delay minimization and throughput maximization.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Sample-Efficient Optimization over Generative Priors via Coarse Learnability
arXiv:2503.06917v5 Announce Type: replace-cross
Abstract: We study zeroth-order optimization where solutions must minimize a cost $d(s)$ while maintaining high probability under a complex generative prior $L(s)$ (e.g., a parameterized model). This reduces to sampling from a target distribution proportional to $L(s) e^{-T \cdot d(s)}$. Since classical model-based optimization (MBO) lacks finite-sample guarantees for expressive approximate learners, we introduce "coarse learnability", a flexible statistical assumption requiring only that a learned model covers the target's probability mass within a polynomial factor. Leveraging this assumption, we design an iterative MBO algorithm called \alift with a sample correction step that provably approximates the target using only a polynomial number of samples. We apply this framework to globally optimizing non-convex objectives bounded by a quadratic envelope in $R^n$, where we show this assumption is naturally satisfied for a family of "optimistic" posterior distributions. To reach global $\varepsilon$-optimality, this implies a sample complexity of $\widetilde{O}(\log 1/\varepsilon)$, a rate characteristic of optimistic space-partitioning methods. We further justify coarse learnability as an assumption for generative priors theoretically, proving that in simple settings, parametric maximum likelihood estimation and over-smoothed kernel density estimators naturally satisfy it. Finally, one motivation for our framework comes from inference-time alignment. Though our primary contribution pertains to the theoretical foundations of MBO, we provide qualitative evidence that, in simple settings, even primitive LLMs can shift their distributions toward lower-cost regions when fine-tuned with zeroth-order feedback.
Fonte: arXiv stat.ML
MLOps/Systems • Score 85
TRACE: A Metrologically-Grounded Engineering Framework for Trustworthy Agentic AI Systems in Operationally Critical Domains
arXiv:2605.03838v1 Announce Type: new
Abstract: We introduce TRACE, a cross-domain engineering framework for trustworthy agentic AI in operationally critical domains. TRACE combines a four-layer reference architecture with an explicit classical-ML vs. LLM-validator split (L2a/L2b), a stateful orchestration-and-escalation policy (L3), and bounded human supervision (L4); a metrologically grounded trust-metric suite mapped to GUM/VIM/ISO 17025; and a Model-Parsimony principle quantified by the Computational Parsimony Ratio (CPR). Three instantiations--clinical decision support, industrial multi-domain operations, and a judicial AI assistant--transfer the samearchitecture and metrics across principally different governance contexts. The L2a/L2b separation makes the use of large language models a deliberate design decision rather than an architectural default, with parsimony quantified through CPR. TRACE introduces CPR as a first-class design principle in trustworthy-AI engineering.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
Faithful Mobile GUI Agents with Guided Advantage Estimator
arXiv:2605.01208v1 Announce Type: new
Abstract: Vision-language model based graphical user interface (GUI) agents have shown strong interaction capabilities. However, they often behave unfaithfully, relying on memorized shortcuts rather than grounding actions in displayed screen evidence or user instructions. To address this, we propose Faithful-Agent, a faithfulness-first framework that reformulates GUI interaction to prioritize evidence groundedness and internal consistency. Faithful-Agent employs a two-stage pipeline: (i) a faithfulness-oriented SFT stage to instill abstainment behaviors under evidence perturbations; (ii) an RFT stage that further amplifies faithfulness by introducing the guided advantage estimator (GuAE), an anchor-based and variance-adaptive advantage tempering mechanism built upon GRPO. GuAE prevents advantage collapse in low-variance rollout groups under sparse GUI rewards, and with a thought-action consistency reward, Faithful-Agent (Stage II) elevates the Trap SR from 13.88\% to 80.21\% relative to the baseline, while preserving robust general instruction-following performance.
Fonte: arXiv cs.AI
Vision • Score 85
TACO: Trajectory Aligning Cross-view Optimisation
arXiv:2605.03315v1 Announce Type: new
Abstract: Cross-View Geo-localisation (CVGL) matches ground imagery against satellite tiles to give absolute position fixes, an alternative to GNSS where signals are occluded, jammed, or spoofed. Recent fine-grained CVGL methods regress sub-tile metric pose, but have only been evaluated as one-shot localisers, never as the primary fix in a live pipeline. Inertial sensing provides high-rate relative motion, but accumulates unbounded drift without an absolute anchor. We propose TACO, a tightly-coupled IMU + fine-grained CVGL pipeline that consumes a single GNSS reading at start-up and thereafter operates on onboard sensing alone. A closed-form cross-track error model triggers CVGL before IMU drift exceeds the matcher's capture radius, and a forward-biased five-point multi-crop search keeps inference cost fixed at five forward passes per fix. A yaw-residual gate rejects fixes that disagree with the onboard compass, and an anisotropic body-frame noise model scales each Unscented Kalman Filter update by per-fix confidence. A factor graph with vetted loop closures provides an offline smoothed trajectory. On the KITTI raw dataset, TACO reduces median Absolute Trajectory Error (ATE) from 97.0m (IMU-only) to 16.3m, a 5.9 times reduction, at <0.1 ms per-frame fusion cost and a 5-10% camera duty cycle. Code is available: github.com/tavisshore/TACO.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
AutoRAGTuner: A Declarative Framework for Automatic Optimization of RAG Pipelines
arXiv:2605.02967v1 Announce Type: new
Abstract: Retrieval-Augmented Generation (RAG) enhances LLMs, but performance is highly sensitive to complex architecture designs and hyper-parameter configurations, which currently rely on inefficient manual tuning. We present AutoRAGTuner, a declarative, configuration-driven framework that automates the RAG life cycle: construction, execution,evaluation, and optimization. AutoRAGTuner employs a modular architecture to decouple pipeline stages through a component registration mechanism. To unify heterogeneous data, we introduce the Domain-Element Model (DEM), representing objects as atomic elements with bidirectional pointers to support nodes, edges, and hyperedges. Furthermore, AutoRAGTuner integrates an adaptive Bayesian optimization engine for end-to-end hyper-parameter tuning. Experimental results demonstrate AutoRAGTuner's architectural generality: across diverse RAG pipelines, ranging from vanilla to graph-based, the framework consistently outperforms default baselines. Notably, AutoRAGTuner significantly mitigates engineering overhead, where its declarative configuration language enables a up to 95\% reduction in code churn for architectural adjustments. Overall, AutoRAGTuner provides a systematically optimizable foundation for building evolvable and reusable RAG systems.
Fonte: arXiv cs.LG
RL • Score 85
CoFlow: Coordinated Few-Step Flow for Offline Multi-Agent Decision Making
arXiv:2605.01457v1 Announce Type: new
Abstract: Generative models have emerged as a major paradigm for offline multi-agent reinforcement learning (MARL), but existing approaches require many iterative sampling steps. Recent few-step accelerations either distill a joint teacher into independent students or apply averaged velocities independently per agent, suggesting that few-step inference requires sacrificing inter-agent coordination. We show this trade-off is not necessary: single-pass multi-agent generation can preserve coordination when the velocity field is natively joint-coupled. We propose Coordinated few-step Flow (CoFlow), an architecture that combines Coordinated Velocity Attention (CVA) with Adaptive Coordination Gating. A finite-difference consistency surrogate further replaces memory-prohibitive Jacobian-vector product backpropagation through the averaged velocity field with two stop-gradient forward passes. Across 60 configurations spanning MPE, MA-MuJoCo, and SMAC, CoFlow matches or surpasses Gaussian / value-based, transformer, diffusion, and prior flow baselines on episodic return. Three independent coordination probes confirm that the gains flow through inter-agent coordination rather than per-agent capacity. A denoising-step sweep shows that single-pass inference suffices on every configuration. CoFlow reaches state-of-the-art coordination quality in 1-3 denoising steps under both centralized and decentralized execution. Project page: https://github.com/Guowei-Zou/coflow.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
SciResearcher: Scaling Deep Research Agents for Frontier Scientific Reasoning
arXiv:2605.01489v1 Announce Type: new
Abstract: Frontier scientific reasoning is rapidly emerging as a key foundation for advancing AI agents in automated scientific discovery. Deep research agents offer a promising approach to this challenge. These models develop robust problem-solving capabilities through post-training on information-seeking tasks, which are typically curated via knowledge graph construction or iterative web browsing. However, these strategies face inherent limitations in frontier science, where domain-specific knowledge is scattered across sparse and heterogeneous academic sources, and problem solving requires sophisticated computation and reasoning far beyond factual recall. To bridge this gap, we introduce SciResearcher, a fully automated agentic framework for frontier-science data construction. SciResearcher synthesizes diverse conceptual and computational tasks grounded in academic evidence, while eliciting information acquisition, tool-integrated reasoning, and long-horizon capabilities. Leveraging the curated data for supervised fine-tuning and agentic reinforcement learning, we develop SciResearcher-8B, an agent foundation model that achieves 19.46% on the HLE-Bio/Chem-Gold benchmark, establishing a new state of the art at its parameter scale and surpassing several larger proprietary agents. It further achieves 13-15% absolute gains on SuperGPQA-Hard-Biology and TRQA-Literature benchmarks. Overall, SciResearcher introduces a new paradigm for automated data construction for frontier scientific reasoning and offers a scalable path toward future scientific agents.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Towards Multi-Agent Autonomous Reasoning in Hydrodynamics
arXiv:2605.01102v1 Announce Type: new
Abstract: Single-agent systems (SAS) have become the default pattern for LLM-driven scientific workflows, but routing planning, tool use, and synthesis through a single context window comes with a well-known cost: as tool specifications and observational traces accumulate, the effective context available for each decision shrinks, and end-to-end reliability suffers. We present a multi-agent system (MAS) prototype for hydrodynamics in which specialized agents are coordinated through a Layer Execution Graph (LEG). A planner agent constructs query-specific execution topologies from natural-language routing heuristics that capture domain knowledge without hard-coding it as rigid control logic; specialist agents operate under strict tool allowlists and occupy complementary data-class roles. Between layers, consolidator agents fuse parallel outputs into concise briefs, and a reporter agent synthesizes the final response, while the runtime logs provenance for every tool invocation to support auditability. All benchmarks, ablations, and stress tests use Claude Sonnet~4.6 as the backbone model for both specialist and general-purpose agents. Evaluated on 37 queries spanning six complexity categories, the prototype achieves 93.6% factual precision with a 100% pass rate. Accuracy remains above 90% across runs from single-threaded to five independent parallel tracks, and under simulated loss of individual data sources the system degrades gracefully, still returning substantive partial answers. Together, these results suggest that planner-guided, graph-structured multi-agent orchestration can meaningfully alleviate the context-saturation bottlenecks that constrain monolithic single-agent architectures.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
Algebraic Semantics of Governed Execution: Monoidal Categories, Effect Algebras, and Coterminous Boundaries
arXiv:2605.01032v2 Announce Type: new
Abstract: We present an algebraic semantics for governed execution in which governance is axiomatized, compositional, and coterminous with expressibility. The framework, mechanized in 32 Rocq modules (~12,000 lines, 454 theorems, 0 admitted), is built on interaction trees and parameterized coinduction. A three-axiom GovernanceAlgebra record (safety, transparency, properness) induces a symmetric monoidal category with verified pentagon, triangle, and hexagon coherence, where every tensor composition preserves governance. An algebraic effect system constrains the handler algebra so that only governance-preserving handlers can be constructed in the safe fragment; programs in the empty capability set provably emit only observability directives. Capability-indexed composition bundles programs with machine-checked capability bounds, and a dual guarantee theorem establishes that within_caps and gov_safe hold simultaneously under all composition operators. The capstone result is the coterminous boundary: within our formal model, every program expressible via the four primitive morphism constructors is governed under interpretation, and every governed program is the image of such a program. Turing completeness is preserved inside governance; unmediated I/O is excluded from the governed fragment. Governance denial is modeled as safe coinductive divergence. The governance algebra is parametric: any system instantiating the three axioms inherits all derived properties, including convergence, compositional closure, and goal preservation. Extracted OCaml runs as a NIF in the BEAM runtime, with property-based testing (70,000+ random inputs, zero disagreements) confirming behavioral equivalence between the specification and the runtime interpreter.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Virtual Speech Therapist: A Clinician-in-the-Loop AI Speech Therapy Agent for Personalized and Supervised Therapy
arXiv:2605.01101v1 Announce Type: new
Abstract: This paper develops Virtual Speech Therapist (VST), an intelligent agent-based platform that streamlines stuttering assessment and delivers customized therapy planning through automated and adaptive AI-driven workflows. VST integrates state-of-the-art deep learning-based stuttering classification, and multi-agent large language model (LLM) reasoning to support evidence-based clinical decision-making. The VST begins with the acquisition and feature extraction of patient speech samples, followed by robust classification of stuttering types. Building on these outputs, VST initiates an agentic reasoning process in which specialized LLM agents autonomously generate, critique, and iteratively refine individualized therapy plans. A dedicated critic agent evaluates all generated therapy plans to ensure clinical safety, methodological soundness, and alignment with peer-reviewed evidence and established professional guidelines. The resulting output is a comprehensive, patient-specific therapy draft intended for clinician review. Incorporating clinician feedback, the system then produces a finalized therapy plan suitable for patient delivery, thereby maintaining a clinician-in-the-loop paradigm. Experimental evaluation by expert speech therapists confirms that VST consistently generates high-quality, evidence-based therapy recommendations. These findings demonstrate the system's potential to augment clinical workflows, reduce clinician burden, and improve therapeutic outcomes for individuals with speech impairments. An interactive user interface for the proposed system is available online at: https://vocametrix.com/ai/stuttering-therapy-planning-agent , facilitating real-time stuttering assessment and personalized therapy planning.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
Conservative quantum offline model-based optimization
arXiv:2506.19714v2 Announce Type: replace-cross
Abstract: Offline model-based optimization (MBO) refers to the task of optimizing a black-box objective function using only a fixed set of prior input-output data, without any active experimentation. Recent work has introduced quantum extremal learning (QEL), which leverages the expressive power of variational quantum circuits to learn accurate surrogate functions by training on a few data points. However, as widely studied in the classical machine learning literature, predictive models may incorrectly extrapolate objective values in unexplored regions, leading to the selection of overly optimistic solutions. In this paper, we propose integrating QEL with conservative objective models (COM) - a regularization technique aimed at ensuring cautious predictions on out-of-distribution inputs. The resulting hybrid algorithm, COM-QEL, builds on the expressive power of quantum neural networks while safeguarding generalization via conservative modeling. Empirical results on benchmark optimization tasks demonstrate that COM-QEL reliably finds solutions with higher true objective values compared to the original QEL, validating its superiority for offline design problems.
Fonte: arXiv stat.ML
Vision • Score 85
The Detector Teaches Itself: Lightweight Self-Supervised Adaptation for Open-Vocabulary Object Detection
arXiv:2605.03642v1 Announce Type: new
Abstract: Open-vocabulary object detection aims to recognize objects from an open set of categories, which leverages vision-language models (VLMs) pre-trained on large-scale image-text data. The cooperative paradigm combines an object detector with a VLM to achieve zero-shot recognition of novel objects. However, VLMs pre-trained on full images often struggle to capture local object details, limiting their effectiveness when applied to region-level detection. We present Decoupled Adaptivity Training (DAT), a self-supervised fine-tuning approach to improve VLMs for cooperative model-based object detection. Given a cooperative model consists of a closed-set detector and a VLM, we first construct a region-aware pseudo-labeled dataset using a pre-trained closed-set object detector, in which regions corresponding to novel objects may be present but remain unlabeled or mislabeled. We then fine-tune the visual backbone of the VLM in a decoupled manner, which enhances local feature alignment while preserving global semantic knowledge via weight interpolation. DAT is a plug-and-play module that requires no inference overhead and fine-tunes less than 0.8M parameters. Experiments on the COCO and LVIS datasets show that DAT consistently improves detection performance on both novel and known categories, establishing a new state of the art in cooperative open-vocabulary detection.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Mantis: Mamba-native Tuning is Efficient for 3D Point Cloud Foundation Models
arXiv:2605.03438v1 Announce Type: new
Abstract: Pre-trained 3D point cloud foundation models (PFMs) have demonstrated strong transferability across diverse downstream tasks. However, full fine-tuning these models is computationally expensive and storage-intensive. Parameter-efficient fine-tuning (PEFT) offers a promising alternative, but existing PEFT approaches are primarily designed for Transformer-based backbones and rely on token-level prompting or feature transformation. Mamba-based backbones introduce a granularity mismatch between token-level adaptation and state-level sequence dynamics. Consequently, straightforward transfer of existing PEFT approaches to frozen Mamba backbones leads to substantial accuracy degradation and unstable optimization. To address this issue, we propose Mantis, the first Mamba-native PEFT framework for 3D PFMs. Specifically, a State-Aware Adapter (SAA) is introduced to inject lightweight task-conditioned control signals into selective state-space updates, enabling state-level adaptation while keeping the pre-trained backbone frozen. Moreover, different valid point cloud serializations are regularized by Dual-Serialization Consistency Distillation (DSCD), thereby reducing serialization-induced instability. Extensive experiments across multiple benchmarks demonstrate that our Mantis achieves competitive performance with only about 5% trainable parameters. Our code is available at https://github.com/gzhhhhhhh/Mantis.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
VLMaxxing through FrameMogging Training-Free Anti-Recomputation for Video Vision-Language Models
arXiv:2605.03351v1 Announce Type: new
Abstract: Video vision-language models (VLMs) keep paying for visual state the stream already told us was stable. The factory wall did not move, but most VLM pipelines still hand the model dense RGB frames or a fresh prefix again. We study that waste as training-free anti-recomputation: reuse state when validation says it survives, and buy fresh evidence when the scene, query, or cache topology requires it.
The largest measured win is after ingest. On frozen Qwen2.5-VL-7B-Instruct-4bit, adaptive same-video follow-up reuse preserves paired choices and correctness on a 93-query VideoMME breadth setting while reducing follow-up latency by 14.90-35.92x. The first query is still cold; the win starts when later questions reuse the same video state. Stress tests bound the result: repeated-question schedules hold through 50 turns, while dense-answer-anchored prompt variation separates conservative fixed K=1 repair from faster aggressive policies that drift.
Fresh-video pruning is smaller but real. C-VISION skips timed vision-tower work before the first answer is generated. On Gemma 4-E4B-4bit, the clean 32f short cell reaches 1.316x first-query speedup with no paired drift or parse failures on 20 items; Qwen shows the fidelity/speed boundary.
Stage-share ceiling (C-CEILING) is the accounting guardrail: a component speedup becomes an end-to-end speedup only in proportion to the wall-clock share it accelerates, so C-VISION and after-ingest follow-up reuse do not multiply. Candidate C-STREAM remains a native-rate target, not a headline result here. The broader direction is VLM-native media that expose change, motion, uncertainty, object state, sensor time, and active tiles directly, so models do not have to rediscover the world from dense RGB every frame.
Fonte: arXiv cs.CV
Privacy/Security/Fairness • Score 85
A Hierarchical Sampling Framework for bounding the Generalization Error of Federated Learning
arXiv:2605.03499v1 Announce Type: cross
Abstract: We study expected generalization bounds for the Hierarchical Federated Learning (HFL) setup using Wasserstein distance. We introduce a generalized framework in which data is sampled hierarchically, and we model it with a multi-layered tree structure that induces dependencies among the clients' datasets. We derive generalization bounds in terms of Wasserstein distance under the Lipschitz assumption on the loss function, by applying a supersample construction that allows us to measure the sensitivity of the algorithm to the change of a single node in the sampling tree. By leveraging the FL structure, we recover and strictly imply existing state-of-the-art conditional mutual information (CMI) bounds in the case of bounded losses. We also show that our bound can be applied together with Differential Privacy assumptions, to recover generalization bounds based on algorithmic privacy. To assess the tightness of our bounds, we study the Gaussian Location Model (GLM) and show that we recover the actual asymptotic rate of the generalization error.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
An End-to-End Framework for Building Large Language Models for Software Operations
arXiv:2605.02906v1 Announce Type: new
Abstract: In the field of software operations, Large Language Models (LLMs) have attracted increasing attention. However, existing research has not yet achieved efficient and effective end-to-end intelligent operations due to low-quality data, fragmented knowledge and insufficient learning. To explore the potential of LLMs in software operations, we propose OpsLLM, a domain-specific LLM that supports both knowledge-based question answering (QA) and root cause analysis (RCA). Moreover, we disclose the detailed workflow for building LLMs specifically in the software operations domain. First, a Human-in-the-Loop mechanism is introduced to curate highquality data from a large collection of operational raw data and construct a fine-tuning dataset. Then, based on the data, supervised fine-tuning is conducted to achieve a base model. Furthermore, we introduce a domain process reward model (DPRM) during the reinforcement learning stage to optimize the accuracy and reliability of the fine-tuned model on RCA tasks. Experimental results on the tasks with diverse difficulties demonstrate that OpsLLMs effectively learns and aligns with the operational domain knowledge infused, outperforming existing open-source and closed-source LLMs in accuracy with improvements of 0.2%~5.7% on QA tasks and 2.7% ~70.3% on RCA tasks, while exhibiting strong transferability. Moreover, we will open-source three versions of OpsLLM with 7B, 14B and 32B parameters, along with a 15K fine-tuning dataset.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
SCALE-LoRA: Auditing Post-Retrieval LoRA Composition with Residual Merging and View Reliability
arXiv:2605.01429v1 Announce Type: new
Abstract: Libraries of Low-Rank Adaptation (LoRA) adapters are becoming a practical by-product of parameter-efficient adaptation. Once such adapters accumulate, a natural question is no longer how to train one adapter for one task, but how to reuse an open pool of adapters for a new task given only a small support set. Prior work has shown that LoRA modules can be composed at the task level and dynamically selected at the instance level. However, open-pool LoRA reuse is not automatic: retrieving relevant adapters does not guarantee that their parameter updates are compatible, and composing adapters does not guarantee reliable outputs.
We introduce the Sparse-Composition Agreement Layer (SCALE), a post-retrieval audit and composition framework for open-pool LoRA reuse. SCALE contains a deployable 1.0* merge path, Layer-Adaptive Sparse Residual Composition (LASRC), and a higher-cost reliability-analysis layer for multi-view disagreement. LASRC addresses merge interference by preserving a linear anchor while residualizing block-wise adapter update directions. The reliability layer treats disagreement among sparse composition views as an observable uncertainty signal and compares agreement, support-loss proxy selection, and oracle headroom under explicit path cost. In matched FLAN-T5-Large, BIG-Bench Hard (BBH), and 97-LoRA experiments, LASRC gives a directional single-view gain under fixed retrieval, while SCALE-support is reported as a query-label-free 3.0* reliability-analysis variant rather than as a calibrated or throughput-equivalent selector. Protocol-distinct BBH-8 validation shows the same qualitative trend on three decoder-only backbones. Detailed scores, paired audits, and path-cost records are reported in the experimental section.
Fonte: arXiv cs.AI
RL • Score 85
Healthcare AI GYM for Medical Agents
arXiv:2605.02943v1 Announce Type: new
Abstract: Clinical reasoning demands multi-step interactions -- gathering patient history, ordering tests, interpreting results, and making safe treatment decisions -- yet a unified training environment provides the breadth of clinical domains and specialized tools to train generalizable medical AI agents through reinforcement learning remains elusive. We present a comprehensive empirical study of multi-turn agentic RL for medical AI, built on \gym{}, a gymnasium-compatible environment spanning 10 clinical domains with 3.6K+ tasks, 135 domain-specific tools, and a knowledge base of 828K medical passages. Our analysis reveals that agentic multi-turn structure degrades into verbose single-turn monologues, characterized by monotonic length explosion and a simultaneous erosion of tool-use frequency. We characterize how this collapse, alongside distillation instability, stems from the misalignment of sparse terminal rewards with sequential clinical trajectories. We find that vanilla GRPO achieves strong final accuracy on some benchmarks but suffers from training instability, evidenced by significant oscillations in response length and prolonged convergence periods. To improve training efficiency and stability, we propose Turn-level Truncated On-Policy Distillation (TT-OPD), a self-distillation framework where a gradient-free EMA teacher leverages outcome-privileged information to provide dense, outcome-aware KL regularization at every conversation turn. TT-OPD achieves the best performance on 10 of 18 benchmarks with an average +3.9~pp improvement over the non-RL baseline with faster early convergence, controlled response length, and sustained multi-turn tool use.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
DeRelayL: Sustainable Decentralized Relay Learning
arXiv:2605.02935v1 Announce Type: new
Abstract: In the era of big data, large-scale machine learning models have revolutionized various fields, driving significant advancements. However, large-scale model training demands high financial and computational resources, which are only affordable by a few technological giants and well-funded institutions. In this case, common users like mobile users, the real creators of valuable data, are often excluded from fully benefiting due to the barriers, while the current methods for accessing large-scale models either limit user ownership or lack sustainability. This growing gap highlights the urgent need for a collaborative model training approach, allowing common users to train and share models. However, existing collaborative model training paradigms, especially federated learning (FL), primarily focus on data privacy and group-based model aggregation. To this end, this paper intends to address this issue by proposing a novel training paradigm named decentralized relay learning (DeRelayL), a sustainable learning system where permissionless participants can contribute to model training in a relay-like manner and share the model. In detail, this paper presents the architecture and workflow of DeRelayL, designs incentive mechanisms to ensure sustainability, and conducts theoretical analysis and numerical simulations to demonstrate its effectiveness.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
MILD: Mediator Agent System with Bidirectional Perception and Multi-Layered Alignment for Human-Vehicle Collaboration
arXiv:2605.01507v1 Announce Type: new
Abstract: Prior studies report that partial driving automation can increase the cognitive demands on human drivers. This effect largely arises from human drivers' lack of transparent insight into the vehicle's intentions and decision logic, as well as from automated systems' limited awareness of the driver's dynamic state and preferences. This bidirectional misalignment undermines shared situational awareness and exacerbates coordination failures in human-vehicle interaction. To address these limitations, we argue for a paradigm shift that elevates the human role from passive supervisor to active manager. We introduce the Mediator-in-the-Loop-Driving (MILD) system, based on an agentic system architecture to facilitate synergistic human-vehicle collaboration. MILD integrates a perception agent for joint in-cabin and out-of-cabin understanding with a lightweight strategy agent that generates compliant and explainable action suggestions. To ensure these strategies are strictly aligned with safety regulations and human values, we develop Evidence- and Constraint-weighted Policy Optimization (ECPO). ECPO leverages automatic validators to steer the agent toward behaviors that are not only accurate but also structurally complete, substantiated by evidence, and free from constraint violations. Furthermore, a retrieval-augmented generation module dynamically incorporates constraints from traffic regulations, speed recommendations, and driver preferences into the decision loop. Field experiments across three open datasets demonstrate that MILD consistently outperforms baselines in both perception accuracy and strategy quality under auditable offline metrics, and yields higher human-rated policy adequacy, comfort, and explanation than baselines. This work offers a practical pathway for building auditable and aligned agents for human-vehicle collaborative driving.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
eOptShrinkQ: Near-Lossless KV Cache Compression Through Optimal Spectral Denoising and Quantization
arXiv:2605.02905v1 Announce Type: new
Abstract: We show that the key-value (KV) cache in transformer attention heads admits a natural decomposition into a low-rank \emph{shared context} component and a full-rank \emph{per-token} residual, well described by the spiked random matrix model. This observation leads to eOptShrinkQ, a two-stage compression pipeline: optimal singular value shrinkage (eOptShrink) automatically extracts the shared structure, and the residual -- which satisfies the \emph{thin shell property} with delocalized coordinates -- is quantized by TurboQuant~\citep{zandieh2025turboquant}, a recently proposed per-vector scalar quantizer with near-optimal distortion guarantees. By restoring the isotropy that scalar quantization assumes, spectral denoising eliminates the need for both outlier handling and dedicated inner product bias correction, freeing those bits for improved reconstruction.
The theoretical grounding in random matrix theory provides three guarantees: automatic rank selection via the BBP phase transition, provably near-zero inner product bias on the residual, and coordinate delocalization ensuring near-optimal quantization distortion. Experimentally, we validate eOptShrinkQ on Llama-3.1-8B and Ministral-8B across three levels: per-head MSE and inner product fidelity, where eOptShrinkQ saves nearly one bit per entry over TurboQuant at equivalent quality; end-to-end on LongBench (16 tasks), where eOptShrinkQ at $\sim$2.2 bits per entry outperforms TurboQuant at 3.0 bits; and multi-needle retrieval, where eOptShrinkQ at 2.2 bits closely matches or exceeds uncompressed FP16, suggesting that spectral denoising can act as a beneficial regularizer for retrieval-intensive tasks.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Tempered Guided Diffusion
arXiv:2605.03712v1 Announce Type: new
Abstract: Training-free conditional diffusion provides a flexible alternative to task-specific conditional model training, but existing samplers often allocate computation inefficiently: independent guided trajectories can vary widely in quality, and additional function evaluations along a single trajectory may not recover from poor early decisions. We propose Tempered Guided Diffusion (TGD), an annealed sequential Monte Carlo framework for training-free conditional sampling with diffusion priors. TGD targets tempered posterior distributions over the clean signal, using noisy diffusion states only as auxiliary variables for proposing reconstructions and propagating particles. Particles are reweighted by incremental likelihood ratios, resampled, and propagated across noise levels, concentrating computation on trajectories plausible under both the prior and observation. Under idealized exact-reconstruction assumptions, full TGD yields a consistent particle approximation to the posterior as the number of particles grows. For expensive reconstruction tasks, Accelerated TGD (A-TGD) retains early particle exploration but prunes to a single high-likelihood trajectory partway through sampling. Experiments on a controlled two-dimensional inverse problem and image inverse problems show improved posterior approximation and favorable wall-clock speed-quality tradeoffs over independent multi-trajectory baselines.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Gated Subspace Inference for Transformer Acceleration
arXiv:2605.03109v1 Announce Type: new
Abstract: A method is presented for accelerating inference in transformer language models by exploiting the low effective rank of the token activation manifold at each layer. The method decomposes each activation vector into a subspace component and a residual, computes the linear-layer output on the subspace component via a cached low-rank weight image at reduced memory bandwidth, and applies a per-token gate that determines whether the residual correction is computed or skipped. The gate ensures that the output distribution is preserved to within a controllable tolerance. Validation on three model families (GPT-2 124M, GPT-J 6B, OPT 6.7B) on AMD MI300X demonstrates effective speedups of 3.0x to 10.5x on linear-layer weight reads with perplexity ratios below 1.00 and top-1 token agreement above 98%. The method requires no retraining, no architectural modification, and no approximation of the attention mechanism. At the operating point (k = 256, {\epsilon} = 0.05) on GPT-J 14 6B, the accelerated model produces character-for-character identical output to the baseline.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Conditional Diffusion Sampling
arXiv:2605.04013v1 Announce Type: new
Abstract: Sampling from unnormalized multimodal distributions with limited density evaluations remains a fundamental challenge in machine learning and natural sciences. Successful approaches construct a bridge between a tractable reference and the target distribution. Parallel Tempering (PT) serves as the gold standard, while recent diffusion-based approaches offer a continuous alternative at the cost of neural training. In this work, we introduce Conditional Diffusion Sampling (CDS), a framework that combines these two paradigms. To this end, we derive Conditional Interpolants, a class of stochastic processes whose transport dynamics are governed by an exact, closed-form stochastic differential equation (SDE), requiring no neural approximation. Although these dynamics require sampling from a non-trivial initialization distribution, we show both theoretically and empirically that the cost of this initialization diminishes for sufficiently short diffusion times. CDS leverages this by a two-stage procedure: (1) PT is used to efficiently sample the initial distribution, and then (2) samples are transported via the transport SDE. This combination couples the robust global exploration of PT with efficient local transport. Experiments suggest that CDS has the potential to achieve a superior trade-off between sample quality and density evaluation cost compared to state-of-the-art samplers.
Fonte: arXiv stat.ML
RL • Score 85
GRPO-TTA: Test-Time Visual Tuning for Vision-Language Models via GRPO-Driven Reinforcement Learning
arXiv:2605.03403v1 Announce Type: new
Abstract: Group Relative Policy Optimization (GRPO) has recently shown strong performance in post-training large language models and vision-language models. It raises a question of whether the GRPO also significantly promotes the test-time adaptation (TTA) of vision language models. In this paper, we propose Group Relative Policy Optimization for Test-Time Adaptation (GRPO-TTA), which adapts GRPO to the TTA setting by reformulating class-specific prompt prediction as a group-wise policy optimization problem. Specifically, we construct output groups by sampling top-K class candidates from CLIP similarity distributions, enabling probability-driven optimization without access to ground-truth labels. Moreover, we design reward functions tailored to test-time adaptation, including alignment rewards and dispersion rewards, to guide effective visual encoder tuning. Extensive experiments across diverse benchmarks demonstrate that GRPO-TTA consistently outperforms existing test-time adaptation methods, with notably larger performance gains under natural distribution shifts.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
PRISM-CTG: A Foundation Model for Cardiotocography Analysis with Multi-View SSL
arXiv:2605.02917v1 Announce Type: new
Abstract: Supervised deep learning models for automated CTG analysis are typically constrained by narrowly curated labelled datasets and limited patient cohorts, leaving substantial volumes of physiologically informative clinical recordings untapped. To address this limitation, we propose Physiology-aware Representation Learning via Integrated Self-supervision and Metadata for CTG (PRISM-CTG), a clinically grounded self-supervised foundation model (FM) for CTG that leverages large-scale unlabelled recordings to learn transferable domain-level representations. PRISM-CTG is pretrained using a multi-view self-supervised framework that jointly optimises 3 complementary pretext objectives: random-projected guided masked signal reconstruction, clinical variable prediction, and feature classification. Each objective is associated with a dedicated task-specific token, enabling specialised representation learning, while controlled cross-attention facilitates information exchange across clinical context. By reframing patient metadata and domain knowledge, which are often underutilised in conventional training as prediction targets, Prism-CTG transforms readily available clinical information into additional supervisory targets that guide clinically meaningful representation learning. Extensive experiments across 7 downstream CTG tasks in both antepartum and intrapartum domains demonstrated that PRISM-CTG consistently outperforms in-domain and SSL baselines. Notably, PRISM-CTG demonstrated strong generalisation under external validation on 2 datasets, while achieving comparable performance to studies trained on substantially larger, privately labelled datasets. To our knowledge, this is the first study to introduce large-scale FM for CTG that learns domain-level representations.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Effect-Transparent Governance for AI Workflow Architectures: Semantic Preservation, Expressive Minimality, and Decidability Boundaries
arXiv:2605.01030v2 Announce Type: new
Abstract: We present a machine-checked formalization of structurally governed AI workflow architectures and prove that effect-level governance can be imposed without reducing internal computational expressivity. Using Interaction Trees in Rocq 8.19, we define a governance operator G that mediates all effectful directives, including memory access, external calls, and oracle (LLM) queries. Our development compiles with 0 admitted lemmas and consists of 36 modules, ~12,000 lines of Rocq, and 454 theorems. We establishseven properties: (P1) governed Turing completeness, (P2) governed oracle expressivity, (P3) a decidability boundary in which governance predicates are total and closed under Boolean composition while semantic program properties remain non-trivial and undecidable by governance, (P4) goal preservation for permitted executions, (P5) expressive minimality of primitive capabilities (compute, memory, reasoning, external call, observability), (P6) subsumption asymmetry showing structural governance strictly subsumes content-level filtering, and (P7) semantic transparency: on all executions where governance permits, the governed interpretation is observationally equivalent (modulo governance-only events) to the ungoverned interpretation. Together, these results show that governance and computational expressivity are orthogonal dimensions: governance constrains the effect boundary of programs while remaining semantically transparent to internal computation.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
2026 Roadmap on Artificial Intelligence and Machine Learning for Smart Manufacturing
arXiv:2605.00839v1 Announce Type: new
Abstract: The evolution of artificial intelligence (AI) and machine learning (ML) is reshaping smart manufacturing by providing new capabilities for efficiency, adaptability, and autonomy across industrial value chains. However, the deployment of AI and ML in industrial settings still faces critical challenges, including the complexity of industrial big data, effective data management, integration with heterogeneous sensing and control systems, and the demand for trustworthy, explainable, and reliable operation in high-stakes industrial environments. In this roadmap, we present a comprehensive perspective on the foundations, applications, and emerging directions of AI and ML in smart manufacturing. It is structured in three parts. The first highlights the foundations and trends that frame the evolution of AI in smart manufacturing. The second focuses on key topics where AI is already enabling advances, including industrial big data analytics, advanced sensing and perception, autonomous systems, additive and laser-based manufacturing, digital twins, robotics, supply chain and logistics optimization, and sustainable manufacturing. The third section explores non-traditional ML approaches that are opening new frontiers, such as physics-informed AI, generative AI, semantic AI, advanced digital twins, explainable AI, RAMS, data-centric metrology, LLMs, and foundation models for highly connected and complex manufacturing systems. By identifying both opportunities and remaining barriers across these areas, this roadmap outlines the advances needed in methods, integration strategies, and industrial adoption. We hope this roadmap will serve as a guide for researchers, engineers, and practitioners to accelerate innovation, align academic and industrial priorities, and ensure that AI-driven smart manufacturing delivers reliable, sustainable, and scalable impact for the future of manufacturing ecosystems.
Fonte: arXiv cs.AI
Evaluation/Benchmarks • Score 85
Evaluating Agentic AI in the Wild: Failure Modes, Drift Patterns, and a Production Evaluation Framework
arXiv:2605.01604v1 Announce Type: new
Abstract: Existing evaluation frameworks for large language models -- including HELM, MT-Bench, AgentBench, and BIG-bench -- are designed for controlled, single-session, lab-scale settings. They do not address the evaluation challenges that emerge when agentic AI systems operate continuously in production: compounding decision errors, tool failure cascades, non-deterministic output drift, and the absence of ground truth for long-horizon tasks. This paper makes three contributions. First, we present a taxonomy of seven failure modes unique to production agentic systems, each grounded in observations from systems operating at billion-event scale. Second, we demonstrate empirically where standard metrics -- ROUGE, BERTScore, accuracy/AUC, and the agentic benchmarks above -- fail to detect each failure mode. Third, we propose PAEF (Production Agentic Evaluation Framework), a five-dimension evaluation framework with an open-source reference implementation, designed for continuous evaluation on production traffic rather than episodic benchmark runs. Our analysis shows that standard metrics fail to detect four of the seven failure modes entirely and detect three others only after a lag of multiple evaluation cycles.
Fonte: arXiv cs.AI
RL • Score 85
Learning to Race in Minutes: Infoprop Dyna on the Mini Wheelbot
arXiv:2605.01096v1 Announce Type: new
Abstract: Reinforcement Learning (RL) has the potential to enable robots with fast, nonlinear, and unstable dynamics to reach the limits of their performance. However, most recent advances rely on carefully designed physics-based simulators and domain randomization to achieve successful sim-to-real transfer within reasonable wall-clock time. In this work, we bypass the need for such simulators and demonstrate that Infoprop Dyna, a state-of-the-art uncertainty-aware model-based reinforcement learning (MBRL) framework, can enable robots to learn directly from real-world interactions. Using Infoprop Dyna, the Mini Wheelbot, an underactuated unicycle robot, learns to race around a track within 11 minutes of real-world experience.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
GIFT: Guided Fine-Tuning and Transfer for Enhancing Instruction-Tuned Language Models
arXiv:2605.01256v1 Announce Type: new
Abstract: A promising paradigm for adapting instruction-tuned language models is to learn task-specific updates on a pretrained base model and subsequently merge them into the instruction-tuned model. However, existing approaches typically treat the instruction-tuned model as a passive target that is only involved at the final merging stage, without guiding the training process. We propose GIFT (Guided Fine-Tuning and Transfer), a simple and efficient framework that incorporates guidance from the instruction model into task adaptation. GIFT fine-tunes a low-rank adapter on the pretrained base model using confidence signals derived from the instruction-tuned model. The learned adapter is then merged into the instruction-tuned model, yielding task-specialized models that preserve general instruction-following behavior. We evaluate GIFT on mathematical and knowledge-intensive benchmarks across multiple model families and scales. Results show that GIFT consistently outperforms direct fine-tuning and representative transfer-based baselines, while maintaining robust generalization and favorable test-time scaling behavior.
Fonte: arXiv cs.CL
Theory/Optimization • Score 85
Stable GFlowNets with Probabilistic Guarantees
arXiv:2605.01729v1 Announce Type: cross
Abstract: Generative Flow Networks (GFlowNets) learn to sample states proportional to an unnormalized reward. Despite their theoretical promise, practical training is often unstable, exhibiting severe loss spikes and mode collapse. To tackle this, we first assess the sensitivity of GFlowNet objectives, demonstrating that a small Total Variation (TV) distance between the learned and target distributions does not preclude unbounded training loss. Motivated by this mismatch, we establish converse guarantees by deriving loss-to-TV bounds that certify global fidelity from bounded trajectory balance losses. Lastly, we propose Stable GFlowNets, an algorithm that leverages our theoretical results to stabilize training, and empirically demonstrate improved training behavior and superior distributional fidelity.
Fonte: arXiv stat.ML
MLOps/Systems • Score 85
LEAP: Layer-wise Exit-Aware Pretraining for Efficient Transformer Inference
arXiv:2605.01058v1 Announce Type: new
Abstract: Layer-aligned distillation and convergence-based early exit represent two predominant computational efficiency paradigms for transformer inference; yet we establish that they exhibit systematic incompatibility under standard deployment conditions for convergence-based early exit. Distillation objectives that align intermediate student layers to teacher representations suppress the representational convergence that early-exit mechanisms exploit, rendering such mechanisms ineffective on distilled models. We introduce LEAP (Layer-wise Exit-Aware Pretraining), an auxiliary training objective that reconciles this incompatibility. LEAP requires no architectural modifications; it augments standard distillation with a single constraint ensuring intermediate layers approximate final-layer representations. LEAP-MiniLM achieves 1.61$\times$ measured wall-clock speedup (batch=1, NVIDIA L4) at $\theta$=0.95, with 91.9% of samples exiting by layer 7 and 1.80$\times$ theoretical layer reduction, where standard distilled models achieve zero effective speedup. We validate across sentence similarity (STS-B: 0.760 $\pm$ 0.006) and retrieval benchmarks (BEIR), providing operational guidance including latency measurements, decision thresholds, and deployment criteria.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
CADFit: Precise Mesh-to-CAD Program Generation with Hybrid Optimization
arXiv:2605.01171v1 Announce Type: new
Abstract: Despite recent progress, recovering parametric CAD construction sequences from geometric input, such as meshes or point clouds, is a key challenge for design and manufacturing, as existing CAD reconstruction and generation methods are largely restricted to difficult-to-edit formats like meshes or Breps or editable simple sketch-and-extrude pipelines and low-complexity datasets. We introduce CADFit, a hybrid optimization-based CAD reconstruction framework that recovers complex, editable CAD construction sequences from meshes by incrementally fitting and validating parametric operations using geometric feedback. Our approach is distinguished by formulating reconstruction as an IoU-driven optimization over structured CAD programs and supporting a rich set of operations, including extrusions, revolutions, fillets, and chamfers. Experiments on multiple CAD benchmarks show that CADFit outperforms state-of-the-art mesh-to-CAD methods in volumetric Intersection-over-Union and Chamfer Distance, while substantially reducing the Invalid Ratio of reconstructed CAD programs, particularly for complex designs. We further present a multimodal pipeline that enables end-to-end reconstruction of CAD construction sequences from images by combining image-based geometry reconstruction with CADFit. By enabling accurate reconstruction of higher-complexity CAD models, CADFit provides a practical foundation for generating richer datasets and advancing future learning-based approaches to CAD reverse engineering.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
Fast Log-Domain Sinkhorn Optimal Transport with Warp-Level GPU Reductions
arXiv:2605.00837v1 Announce Type: new
Abstract: Entropic regularized optimal transport (OT) via the Sinkhorn algorithm has become a fundamental tool in machine learning, yet existing implementations either suffer from numerical instability for small regularization parameters or incur significant overhead from deep learning frameworks. We present FastSinkhorn, a lightweight, native CUDA implementation of the log-domain Sinkhorn algorithm that combines warp-level shuffle reductions with shared-memory tiling to achieve high GPU utilization without sacrificing numerical stability. Our solver operates entirely in the log-domain, enabling robust computation for regularization parameters as small as epsilon = 10^{-4} where standard-domain methods fail. On dense OT problems with n = m = 8192, our implementation achieves 12x speedup over the widely-used POT library and 5.9x speedup over GPU-accelerated PyTorch baselines, while consuming only 256 MB of GPU memory. We validate our solver on image color transfer, 3D point cloud matching, and convergence analysis, demonstrating that native CUDA kernels with careful numerical treatment provide a practical and efficient foundation for large-scale optimal transport computation.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Diffusion Operator Geometry of Feedforward Representations
arXiv:2605.01107v1 Announce Type: new
Abstract: Neural networks transform data through learned representations whose geometry affects separation, contraction, and generalization. Recent work studies this geometry using discrete curvature on neighborhood graphs, suggesting Ricci-flow-like behavior across layers. We develop a smooth operator-theoretic alternative for feedforward representation snapshots. Each feature cloud induces a Gaussian-kernel diffusion Markov operator, and transport, spectral, label-boundary, and local-scale observables are derived from this single object via Bakry-Emery $\Gamma$-calculus. In a balanced Gaussian class-conditional snapshot model with shared covariance, the population operator has closed-form class affinities, leakage, and coarse spectra, all controlled by pairwise regularized Mahalanobis separations $c_\varepsilon^{(a,b)}$. We also prove that the resulting operator observables vary smoothly under feature perturbations, while hard neighborhood-graph diagnostics can change discontinuously. Synthetic experiments validate the closed-form Gaussian bridge, while learned MNIST experiments show that the same operator observables track training, width, and perturbation stability. Together, these results give a stable operator-geometric framework for analyzing feedforward representation geometry.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
When To Adapt? Adapting the Model or Data in Federated Medical Imaging
arXiv:2605.00892v1 Announce Type: new
Abstract: Federated learning enables collaborative model training across medical institutions without sharing raw data, but its performance is often limited by domain heterogeneity across clients. Existing approaches to address this challenge fall into two main paradigms: model-side personalization, which adapts model parameters to each client, and data-side harmonization, which reduces inter-client variation at the input level. Despite their widespread use, these strategies have not been systematically compared. In this work, we conduct a comprehensive study across six medical imaging settings-colon polyp, skin lesion, and breast tumor segmentation, and tuberculosis CXR, brain tumor, and breast tumor classification-covering diverse types of domain shift. We evaluate a broad set of state-of-the-art harmonization and personalization methods under a unified framework. Our results reveal a conditional trade-off driven by the nature of heterogeneity: harmonization is more effective when variation is primarily appearance-based (e.g., CXR classification), while personalization performs better when differences are structural (e.g., colon polyp segmentation). When inter-client variation is limited, both strategies perform similarly. These findings demonstrate that the effectiveness of adaptation in federated medical imaging depends on the type and magnitude of domain shift rather than the strategy alone. We provide practical guidelines for selecting between harmonization and personalization and highlight directions for future hybrid approaches that combine both paradigms. Code is available at https://github.com/ChamaniS/WhenToAdapt.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
EventADL: Open-Box Anomaly Detection and Localization Framework for Events in Cloud-Based Service Systems
arXiv:2605.00936v1 Announce Type: new
Abstract: Anomaly detection and localization (ADL) is critical for maintaining reliability and availability in cloud systems. Recent ADL developments focus on metric and log data, leaving event data unexplored. To address this gap, we propose EventADL, the first open-box event-based ADL framework for cloud-based service systems. To motivate the design of our framework, we conduct a systematic analysis on 520 real-world incidents, and provide insights into how anomalies and their root causes manifest through event data. EventADL has three phases: offline training, online anomaly detection, and root cause localization. During the training phase, EventADL first learns Event Semantic Patterns (ESPs), which capture normal interactions between system entities using historical event data, and then learns Event Frequency Patterns (EFPs), which capture the normal frequency of known ESPs. In the online anomaly detection phase, any data in the event stream that deviates significantly from either pattern is identified as anomalous. For localization, EventADL constructs an Intervention Graph that models the relationships between recent system interactions and the detected anomalies for automatic root cause localization. The framework is designed to operate efficiently with unlabeled data and to produce interpretable anomalies with their corresponding root causes. Our evaluation on three real cloud service systems and two real-world incidents demonstrates that EventADL outperforms existing methods, achieving F1-scores of at least 90% for anomaly detection and 100% top-3 accuracy in root cause localization.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Phase-map synthesis from magnitude-only MR images using conditional score-based diffusion models with application in training of accelerated MRI reconstruction models
arXiv:2605.01185v1 Announce Type: new
Abstract: Accelerated magnetic resonance imaging (MRI) enabled by the training of deep learning (DL)-based image recon. models requires large and diverse raw k-space datasets. In most clinical MRI applications, due to storage and patient privacy concerns, raw k-space data is discarded and magnitude-only images are the only component saved. Consequently, a large portion of the DL-based MRI recon. literature has either relied on small training datasets or has used one of the few available open-source k-space datasets. At the same time, the growing number of anonymized magnitude-only image registries/databases motivates the development of techniques that can use them as training datasets for generalizable DL-based recon. models. Here we propose to address this challenge by employing a generative approach based on conditional score-based diffusion models (SBDMs): given a magnitude-only MR image, it synthesizes a phase map (in the image domain) that realistically corresponds to the magnitude-only image. We evaluate its generative capabilities in a downstream DL-based recon. task whereby a large k-space dataset is generated by combining the SBDM-synthesized phase-maps and the corresponding magnitude-only images, and this k-space dataset is then used to train a DL model for accelerated MRI recon. We compare the performance of the resulting DL model versus those trained according to (a) a naive approach that uses smooth phase, (b) a k-space training dataset generated using synthesized phase maps derived from a generative adversarial network, and (c) the ground truth k-space data. Our results suggest that the DL model trained from SBDM-synthesized k-space data outperforms the other approaches in terms of quantitative metrics as well as qualitatively observed recon. fidelity, i.e., whether the reconstructed images include erroneous or hallucinated features that could adversely impact diagnostic accuracy.
Fonte: arXiv cs.CV
RL • Score 85
Learn where to Click from Yourself: On-Policy Self-Distillation for GUI Grounding
arXiv:2605.00642v1 Announce Type: new
Abstract: Graphical User Interface (GUI) grounding maps natural language instructions to the visual coordinates of target elements and serves as a core capability for autonomous GUI agents. Recent reinforcement learning methods (e.g., GRPO) have achieved strong performance, but they rely on expensive multiple rollouts and suffer from sparse signals on hard samples. These limitations make on-policy self-distillation (OPSD), which provides dense token-level supervision from a single rollout, a promising alternative. However, its applicability to GUI grounding remains unexplored. In this paper, we present GUI-SD, the first OPSD framework tailored for GUI grounding. First, it constructs a visually enriched privileged context for the teacher using a target bounding box and a Gaussian soft mask, providing informative guidance without leaking exact coordinates. Second, it employs entropy-guided distillation, which adaptively weights tokens based on digit significance and teacher confidence, concentrating optimization on the most impactful and reliable positions. Extensive experiments on six representative GUI grounding benchmarks show that GUI-SD consistently outperforms GRPO-based methods and naive OPSD in both accuracy and training efficiency. Code and training data are available at https://zhangyan-ucas.github.io/GUI-SD/.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
SiriusHelper: An LLM Agent-Based Operations Assistant for Big Data Platforms
arXiv:2605.00043v1 Announce Type: cross
Abstract: Big data platforms are widely used in modern enterprises, and an in-production intelligent assistant is increasingly important to help users quickly find actionable guidance and reduce operational burden. While recent LLM+RAG assistants provide a natural interface, they face practical challenges in real deployments: limited scenario coverage across both general consultation and domain-specific troubleshooting workflows, inefficient knowledge access due to inadequate multi-hop retrieval and flat knowledge organization, and high maintenance cost because escalated tickets are unstructured and hard to convert into assistant improvements and reusable SOPs.
In this paper, we present SiriusHelper, a deployed intelligent assistant for big data platforms. SiriusHelper serves as a unified online assistant that automatically identifies user intent and routes queries to the right handling path, including dedicated expert workflows for specialized scenarios (e.g., SQL execution diagnosis). To support complex troubleshooting, SiriusHelper combines a DeepSearch-driven mechanism with a priority-based hierarchical knowledge base to enable multi-hop retrieval without context overload, thus improving answer reliability and latency. To reduce expert overhead, SiriusHelper further introduces automated ticket understanding and SOP distillation: it diagnoses the assistant failure reason (e.g., missing knowledge or wrong routing) and extracts domain-specific SOPs to continuously enrich the knowledge base. Experiments and online deployment on Tencent Big Data platform show that SiriusHelper outperforms representative alternatives and reduces online ticket volume by 20.8\%.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Hierarchical Federated Learning for Networked AI: From Communication Saving to Architecture-Aware Design
arXiv:2605.00931v1 Announce Type: new
Abstract: Federated learning (FL) is fundamentally a distributed optimization problem executed by communicating agents with local data, local computation, and partial system visibility. Once FL is viewed through that lens, hierarchy is not merely a scalability mechanism. It becomes the natural place to rethink how distributed optimization should be organized over real multi-tier networks. This article argues that hierarchical federated learning (HFL) should move beyond its common framing as a communication-saving protocol and instead be viewed as an architecture-aware design framework for networked AI. The framework is organized around three coupled design axes: architectural parameters, layer-wise optimization decomposition, and layer-wise communication realization. The first axis determines the coordination geometry of learning through hierarchy depth, layer asymmetry, and layered connectivity. The second determines how the global FL objective is decomposed across layers and highlights modular multi-layer optimization as a major opportunity beyond one dominant method everywhere. The third determines how the distributed optimization is physically realized under heterogeneous communication regimes, from interference-limited lower tiers to reliable upper tiers. A central message is that, in HFL, convergence becomes architecture-dependent: it is directly shaped by the chosen hierarchy, the assigned optimization roles, and the communication mechanisms that connect them. We develop this viewpoint using large-scale wireless edge intelligence as a flagship networked AI setting, then provide a comparative perspective on flat FL, two-tier HFL, and deep HFL together with a regime-oriented design map. The resulting perspective positions HFL as a practical methodology for designing future networked AI systems.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
CombinationTS: A Modular Framework for Understanding Time-Series Forecasting Models
arXiv:2605.01231v1 Announce Type: new
Abstract: Recent progress in time-series forecasting has led to rapidly increasing architectural complexity, yet many reported State-of-the-Art gains are statistically fragile or misattributed. We argue that progress requires a shift from model selection to modular attribution, identifying which components truly drive performance. We propose CombinationTS, a self-contained probabilistic evaluation framework that decomposes forecasting models into orthogonal modules--Input Transformation, Embedding, Encoder, Decoder, and Output Transformation--and evaluates them under a shared evaluation condition space. By quantifying each component via marginalized performance ($\mu$) and stability ($\sigma$), CombinationTS enables robust attribution beyond fragile point estimates. Through large-scale paired evaluation, we uncover the Identity Paradox: once the data view (Embedding) is well-designed, a parameter-free Identity Encoder often matches or outperforms complex backbones. We further show that explicit structural priors introduced via Input Transformations yield a more favorable performance-stability trade-off than increasing Encoder complexity, establishing a principled baseline for architectural necessity.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Active multiple matrix completion with adaptive confidence sets
arXiv:2605.02458v1 Announce Type: new
Abstract: In this work, we formulate a new multi-task active learning setting in which the learner's goal is to solve multiple matrix completion problems simultaneously. At each round, the learner can choose from which matrix it receives a sample from an entry drawn uniformly at random. Our main practical motivation is market segmentation, where the matrices represent different regions with different preferences of the customers. The challenge in this setting is that each of the matrices can be of a different size and also of a different rank which is unknown. We provide and analyze a new algorithm, MAlocate that is able to adapt to the unknown ranks of the different matrices. We then give a lower-bound showing that our strategy is minimax-optimal and demonstrate its performance with synthetic experiments.
Fonte: arXiv stat.ML
MLOps/Systems • Score 85
Neighbor2Inverse: Self-Supervised Denoising for Low-Dose Region-of-Interest Phase Contrast CT
arXiv:2605.01075v1 Announce Type: new
Abstract: Propagation-based X-ray phase-contrast imaging (PBI) enables high-contrast visualization of lung structures and holds strong medical potential. However, safe translation to the clinic will require a substantial radiation dose reduction, which inevitably increases image noise. Supervised convolutional-neural-network-based denoising can restore image quality but depends on paired low- and high-dose datasets, which are rarely available in practice. Self-supervised methods avoid this limitation, yet most are not well adapted to the inverse problem of PBI computed tomography (CT). We introduce Neighbor2Inverse, a self-supervised denoising framework designed for low-dose PBI-CT that generalizes to clinical CT. Building on the Neighbor2Neighbor principle, each noisy projection is subsampled into two variants that preserve structural information but contain independent noise realizations. These are reconstructed separately, and the resulting pairs are used to train a denoising network directly in the image domain. We benchmark the proposed method against established analytical and self-supervised denoising approaches. In region-of-interest PBI CT experiments, Neighbor2Inverse achieves superior noise suppression while preserving fine structural details, as demonstrated by improved contrast-to-noise ratio, spatial resolution, and composite image quality metrics. Competitive performance is also observed on clinical CT data under simulated low-dose conditions.
This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible.
Code, data, and interactive figures are available at https://github.com/J-3TO/Neighbor2Inverse.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
AgentReputation: A Decentralized Agentic AI Reputation Framework
arXiv:2605.00073v1 Announce Type: new
Abstract: Decentralized, agentic AI marketplaces are rapidly emerging to support software engineering tasks such as debugging, patch generation, and security auditing, often operating without centralized oversight. However, existing reputation mechanisms fail in this setting for three fundamental reasons: agents can strategically optimize against evaluation procedures; demonstrated competence does not reliably transfer across heterogeneous task contexts; and verification rigor varies widely, from lightweight automated checks to costly expert review. Current approaches to reputation drawing on federated learning, blockchain-based AI platforms, and large language model safety research are unable to address these challenges in combination. We therefore propose \textbf{AgentReputation}, a decentralized, three-layer reputation framework for agentic AI systems. The framework separates task execution, reputation services, and tamper-proof persistence to both leverage their respective strengths and enable independent evolution. The framework introduces explicit verification regimes linked to agent reputation metadata, as well as context-conditioned reputation cards that prevent reputation conflation across domains and task types. In addition, AgentReputation provides a decision-facing policy engine that supports resource allocation, access control, and adaptive verification escalation based on risk and uncertainty. Building on this framework, we outline several future research directions, including the development of verification ontologies, methods for quantifying verification strength, privacy-preserving evidence mechanisms, cold-start reputation bootstrapping, and defenses against adversarial manipulation.
Fonte: arXiv cs.AI
Privacy/Security/Fairness • Score 85
Ambient Persuasion in a Deployed AI Agent: Unauthorized Escalation Following Routine Non-Adversarial Content Exposure
arXiv:2605.00055v1 Announce Type: cross
Abstract: We report a safety incident in a deployed multi-agent research system in which a primary AI agent installed 107 unauthorized software components, overwrote a system registry, overrode a prior negative decision from an oversight agent, and escalated through increasingly privileged operations up to an attempted system administrator command. The incident was preceded not by an adversarial attack but by routine content: a forwarded technology article written for human developers and shared by the principal investigator for discussion. The agent operated in a permissive environment, with unrestricted shell access, soft behavioral guidelines containing genuinely conflicting instructions, and no machine-enforced installation policy, and had recommended installing the same tool six hours earlier before being told to stand down. We analyze the behavioral cascade, the control boundaries that failed, and the limitations of multi-agent oversight in detecting and remediating the damage. We use directive weighting error as a descriptive interpretation of the observed failure and ambient persuasion as a provisional analytic label for the broader trigger configuration of non-adversarial environmental content preceding unauthorized agent action. The case highlights ethical and governance implications for deployed agent systems: ambiguous conversational cues are insufficient authorization for consequential actions, prior refusals must persist as enforceable constraints rather than message-level reminders, and oversight mechanisms require systematic post-incident auditing in addition to routine monitoring.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
AgentFloor: How Far Up the tool use Ladder Can Small Open-Weight Models Go?
arXiv:2605.00334v1 Announce Type: new
Abstract: Production agentic systems make many model calls per user request, and most of those calls are short, structured, and routine. This raises a practical routing question that existing evaluations do not directly answer: which parts of an agent workflow truly require large frontier intelligence, and which can be handled by smaller models? We introduce AgentFloor, a deterministic 30-task benchmark organized as a six-tier capability ladder, spanning instruction following, tool use, multi-step coordination, and long-horizon planning under persistent constraints. We evaluate 16 open-weight models, from 0.27B to 32B parameters, alongside GPT-5 across 16,542 scored runs. Our results reveal a clear boundary of model necessity. Small and mid-sized open-weight models are already sufficient for much of the short-horizon, structured tool use work that dominates real agent pipelines, and in aggregate, the strongest open-weight model matches GPT-5 on our benchmark while being substantially cheaper and faster to run. The gap appears most clearly on long-horizon planning tasks that require sustained coordination and reliable constraint tracking over many steps, where frontier models still hold an advantage, though neither side reaches strong reliability. We also find that this boundary is not explained by scale alone: some failures respond to targeted interventions, but the effects are model-specific rather than universal. These findings suggest a practical design principle for agentic systems: use smaller open-weight models for the broad base of routine actions, and reserve large frontier models for the narrower class of tasks that truly demand deeper planning and control. We release the benchmark, harness, sweep configurations, and full run corpus.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
When Less is Enough: Efficient Inference via Collaborative Reasoning
arXiv:2605.01111v1 Announce Type: new
Abstract: In this work, we introduce DUET (Dual-model Efficient Two-stage inference), a collaborative inference framework in which a capable model and a lightweight model work together to solve a task. Relying on a single large model to perform end-to-end reasoning and prediction often incurs substantial inference cost. In contrast, DUET decomposes inference into two stages: the capable model produces a reasoning signal, and the lightweight model interprets this signal to generate the final answer, allowing reasoning-intensive computation to be handled by the capable model while non-reasoning-intensive components are delegated to the lightweight model without sacrificing task performance. To achieve this objective, we propose a length-penalized joint training objective that encourages the capable model to transmit only the information that is sufficient for the lightweight model to solve the task. As a result, DUET maintains strong reasoning performance with substantially lower inference cost than end-to-end inference using a large model alone, saving up to 60% of the large model's output tokens on challenging reasoning benchmarks, including AIME and GPQA.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
CGM-JEPA: Learning Consistent Continuous Glucose Monitor Representations via Predictive Self-Supervised Pretraining
arXiv:2605.00933v1 Announce Type: new
Abstract: Continuous Glucose Monitoring (CGM) can detect early metabolic subphenotypes (insulin resistance, IR; $\beta$-cell dysfunction), but population-scale deployment faces two coupled problems. First, the same physiological state appears through multiple views (CGM time series, venous OGTT, Glucodensity summaries), so single-view representations fail to transfer when deployment shifts the modality or setting. Second, baselines perform inconsistently across these shifts. Both problems point to one remedy: representations that abstract away from any single view to capture higher-level temporal and distributional structure. We propose CGM-JEPA, a self-supervised pretraining framework which predicts masked latent representations rather than raw values, yielding abstraction that transfers across modalities. X-CGM-JEPA adds a masked Glucodensity cross-view objective for complementary distributional information. We pretrain on $\sim$389k unlabeled CGM readings from 228 subjects and evaluate on two clinical cohorts ($N=27$ and $N=17$ public-release subsets) across three regimes (cohort generalization, venous-to-CGM transfer, home CGM) under 20-iteration $\times$ 2-fold cross-validation. X-CGM-JEPA ranks first or second on AUROC for both endpoints across all three regimes while no baseline does, exceeding the strongest baseline by up to $+6.5$ pp in cohort generalization and $+3.6$ pp in venous-to-CGM transfer (paired Wilcoxon, $p<0.001$). Under modality shift, it matches mean AUROC while redistributing toward weaker subgroups (ethnicity AUROC gap shrinks 25-54%); on sparse in-domain venous data, the distributional view lifts label-aware clustering (ARI $+39\%$, NMI $+40\%$). Code and weights: https://github.com/cruiseresearchgroup/CGM-JEPA
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
PhaseNet++: Phase-Aware Frequency-Domain Anomaly Detection for Industrial Control Systems via Phase Coherence Graphs
arXiv:2605.00929v1 Announce Type: new
Abstract: Multivariate time series anomaly detection in ICS has attracted growing attention due to the increasing threat of cyber-physical attacks on critical infrastructure. State-of-the-art methods model inter-sensor relationships from raw time-domain amplitude values, using graph neural networks, Transformers. However, these methods discard the phase spectrum produced by time frequency transformations, We argue that phase information constitutes a complementary and previously overlooked detection modality for ICS anomaly detection. We present PhaseNet++, a frequency-domain autoencoder that operates on the Short-Time Fourier Transform (STFT) of sliding sensor windows, retaining both magnitude and phase spectra. A Phase Coherence Index (PCI), inspired by the Phase Locking Value from neuroscience, summarizes pairwise phase consistency across frequency bins into a continuous adjacency matrix. This matrix guides a graph attention network that propagates information preferentially among phase-synchronized sensors. A sensor-token Transformer encoder captures system-wide structure, and a dual-head decoder reconstructs magnitude and phase jointly via circular and coherence-aware objectives. Evaluated on the Secure Water Treatment (SWaT) benchmark, PhaseNet++ achieves an F1-score of 90.98%, ROC-AUC of 95.66%, and average precision of 91.51%. Ablation studies show that the phase-aware front-end and PCI graph module together add only 264,816 parameters, demonstrating that the phase inductive bias is lightweight. While the absolute F1-score is second best than that of all recent raw-value methods evaluated under different protocols, we position this work as the first systematic study of phase-domain anomaly detection for ICS.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Machine Learning-Augmented Acceleration of Iterative Ptychographic Reconstruction
arXiv:2605.01122v1 Announce Type: new
Abstract: Iterative ptychographic reconstruction algorithms are widely used for coherent diffractive imaging but can exhibit slow convergence under realistic experimental conditions. We propose a machine learning-augmented approach that accelerates iterative ptychographic reconstruction by introducing a learned fast-forward operator applied during reconstruction. Following an initial warm-up using standard iterations, the fast-forward operator advances the reconstruction toward a more converged state, after which conventional iterative updates are resumed. This strategy preserves the physical consistency and flexibility of established ptychographic solvers while reducing the number of iterations required for convergence. The model is trained on diverse ptychographic datasets and evaluated on experimental data acquired in a different year, demonstrating robustness and temporal generalization. Compared with conventional iterative solvers, the machine learning-augmented method achieves comparable reconstruction quality while converging faster in terms of Poisson negative log-likelihood, yielding over a two-fold reduction in wall-clock time. The approach has been integrated into an existing reconstruction pipeline and deployed in production at a synchrotron beamline, demonstrating practicality for real-time experimental operation.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Missingness-aware Data Imputation via AI-powered Bayesian Generative Modeling
arXiv:2605.01676v1 Announce Type: new
Abstract: Missing data imputation remains a fundamental challenge in modern data science, especially when uncertainty quantification is essential. In this work, we propose MissBGM, an AI-powered missing data imputation method via Bayesian generative modeling that bridges the expressive flexibility of neural networks with the statistical rigor of Bayesian inference. Unlike existing methods that often focus on point estimates or treat the missingness mechanism implicitly, MissBGM explicitly and jointly models the data-generating and missingness mechanisms, providing principled posterior uncertainty over imputations rather than a single point estimate. We develop a stochastic optimization framework with alternating updates among missing values, model parameters, and latent variables until convergence. Our theoretical analysis shows that estimates of missing values from MissBGM converge consistently under mild assumptions. Empirically, we demonstrate that MissBGM achieves superior performance over traditional imputers and recent neural network-based methods across extensive experimental settings. These results establish MissBGM as a principled and scalable solution for modern missing data imputation. The code for MissBGM is open sourced at https://github.com/liuq-lab/MissBGM.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Thinking in Text and Images: Interleaved Vision--Language Reasoning Traces for Long-Horizon Robot Manipulation
arXiv:2605.00438v1 Announce Type: new
Abstract: Long-horizon robotic manipulation requires plans that are both logically coherent and geometrically grounded. Existing Vision-Language-Action policies usually hide planning in latent states or expose only one modality: text-only chain-of-thought encodes causal order but misses spatial constraints, while visual prediction provides geometric cues but often remains local and semantically underconstrained. We introduce Interleaved Vision--Language Reasoning (IVLR), a policy framework built around \trace{}, an explicit intermediate representation that alternates textual subgoals with visual keyframes over the full task horizon. At test time, a single native multimodal transformer self-generates this global semantic-geometric trace from the initial observation and instruction, caches it, and conditions a closed-loop action decoder on the trace, original instruction, and current observation. Because standard robot datasets lack such traces, we construct pseudo-supervision by temporally segmenting demonstrations and captioning each stage with a vision-language model. Across simulated benchmarks for long-horizon manipulation and visual distribution shift, \method{} reaches 95.5\% average success on LIBERO, including 92.4\% on LIBERO-Long, and 59.4\% overall success on SimplerEnv-WidowX. Ablations show that both modalities are necessary: without traces, LIBERO-Long success drops to 37.7\%; text-only and vision-only traces reach 62.0\% and 68.4\%, while the full interleaved trace reaches 92.4\%. Stress tests with execution perturbations and masked trace content show moderate degradation, suggesting that the trace can tolerate local corruption and moderate execution drift, but remains limited under stale or incorrect global plans.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
TADI: Tool-Augmented Drilling Intelligence via Agentic LLM Orchestration over Heterogeneous Wellsite Data
arXiv:2605.00060v1 Announce Type: new
Abstract: We present TADI (Tool-Augmented Drilling Intelligence), an agentic AI system that transforms drilling operational data into evidence-based analytical intelligence. Applied to the Equinor Volve Field dataset, TADI integrates 1,759 daily drilling reports, selected WITSML real-time objects, 15,634 production records, formation tops, and perforations into a dual-store architecture: DuckDB for structured queries over 12 tables with 65,447 rows, and ChromaDB for semantic search over 36,709 embedded documents. Twelve domain-specialized tools, orchestrated by a large language model via iterative function calling, support multi-step evidence gathering that cross-references structured drilling measurements with daily report narratives. The system parses all 1,759 DDR XML files with zero errors, handles three incompatible well naming conventions, and is backed by 95 automated tests plus a 130-question stress-question taxonomy spanning six operational categories. We formalize the agent's behavior as a sequential tool-selection problem and propose the Evidence Grounding Score (EGS) as a simple grounding-compliance proxy based on measurements, attributed DDR quotations, and required answer sections. The complete 6,084-line, framework-free implementation is reproducible given the public Volve download and an API key, and the case studies and qualitative ablation analysis suggest that domain-specialized tool design, rather than model scale alone, is the primary driver of analytical quality in technical operations.
Fonte: arXiv cs.AI
Applications • Score 85
Agentic AI for Trip Planning Optimization Application
arXiv:2605.00276v1 Announce Type: new
Abstract: Trip planning for intelligent vehicles increasingly requires selecting optimal routes rather than merely producing feasible itineraries, as interacting factors such as travel time, energy consumption, and traffic conditions directly affect plan quality. Yet existing systems are largely designed for feasibility-oriented planning, and current benchmarks provide only reference answers without ground truth, preventing objective evaluation of optimization performance. In our paper, we address these limitations with an agentic AI framework that enables dynamic refinement through an orchestration agent coordinating specialized agents for traffic, charging, and points of interest, and with the Trip-planning Optimization Problems Dataset, which supplies definitive optimal solutions and category-level task structure for fine-grained analysis. Experiments show that our system achieves 77.4\% accuracy on the TOP Benchmark, significantly outperforming single-agent and workflow-based multi-agent baselines, demonstrating the importance of orchestrated agentic reasoning for robust trip planning optimization.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Selective Correlation Based Knowledge Distillation for Ground Reaction Force Estimation
arXiv:2605.00888v1 Announce Type: new
Abstract: Wearable sensor-based human gait analysis holds great promise in healthcare, rehabilitation, clinical diagnosis and monitoring, and sports activities. Specifically, ground reaction force (GRF) provides essential insights into the body's interaction with the ground during movement and is typically measured using instrumented treadmills equipped with force plates. However, such equipment is expensive and restricted to laboratory environments. To enable a more portable solution, wearable insole sensors have been used to measure GRF. These sensors, however, are prone to noise and external interference, which reduces measurement accuracy. Deep learning methodologies could be adopted to address these issues, but they often require significant computing resources to achieve high accuracy, limiting their applicability for real-time analysis on portable devices. To overcome these limitations, we propose Selective Correlation Based Knowledge Distillation (SCKD) for estimating GRF from data collected by insole sensors. Our proposed method utilizes selected features considering temporal characteristics in the process of extracting correlation maps for knowledge transfer, enhancing interpretability and mitigating issues in high dimensional data processing. We demonstrate the effectiveness of the compact models generated by our distillation framework through comparison with existing methods. Various configurations of teacher-student architectures and training approaches are examined based on multiple evaluation criteria, utilizing data collected at different walking speeds and with different window sizes. Experimental results confirm that our approach outperforms existing methods in estimating GRF from wearable insole sensor data. Therefore, our approach offers a reliable and resource-efficient solution for human gait analysis.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Visual Chart Representations for Cryptocurrency Regime Prediction: A Systematic Deep Learning Study
arXiv:2605.00875v1 Announce Type: new
Abstract: Technical traders have long relied on visual analysis of candlestick charts to identify market patterns and predict price movements. While deep learning has achieved remarkable success in image classification, its application to financial chart images remains underexplored. This paper presents a systematic study comparing different visual representations for cryptocurrency regime prediction. We evaluate three image encoding methods (raw candlestick charts, Gramian Angular Fields, and multi-channel GAF), five chart component configurations, four neural network architectures (CNN, ResNet18, EfficientNet-B0, and Vision Transformer), and the impact of ImageNet transfer learning. Through eight controlled experiments on Bitcoin, Ethereum, and S&P 500 data spanning 2018-2024, we identify optimal configurations for visual regime classification. Our results show that a simple 4-layer CNN on raw candlestick charts achieves 0.892 AUC-ROC, outperforming larger pretrained models. Surprisingly, simpler representations (price-only charts, 128x128 resolution) consistently outperform more complex alternatives. We provide interpretability analysis using GradCAM and demonstrate that transfer learning improves performance by 4-16% despite the domain gap between natural images and financial charts.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
Compliance-Aware Agentic Payments on Stablecoin Rails
arXiv:2605.00071v1 Announce Type: cross
Abstract: Agentic payment systems extend delegated action to financial transfers, but scaling them on stablecoin rails in regulated settings requires safeguards that remain effective when humans are not continuously in the loop. We present a compliance-aware architecture that combines x402-style, signature-based payment authorisation and relayed execution with programmable compliance embedded as an on-chain guardrail via a policy wrapper and policy manager coordinating modular checks. By enforcing compliance at the point of execution, rather than as a separate off-chain workflow, the approach preserves low-friction settlement when conditions are satisfied, records transaction-linked on-chain attestations, and supports structured resolution when requirements are pending.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
Learning Discriminators for Resampling in the Ensemble Gaussian Mixture Filter through a Normalizing Flow Approach
arXiv:2605.01089v1 Announce Type: new
Abstract: The ensemble Gaussian mixture filter (EnGMF) is a powerful, convergent particle filter capable of medium-to-high dimensional non-linear filtering. The EnGMF relies on a resampling step that can generate physically unrealistic posterior samples, that would subsequently produce physically meaningless forecasts. This work introduces the discriminator-informed resampling procedure, that augments the posterior resampling step with a discriminator that accepts or rejects candidate particles based on their physical plausibility. In this work these discriminators are learned through a normalizing flow approach. Numerical experiments on both the Ikeda map and the Lorenz '63 system show that discriminator informed resampling procedure consistently reduces error relative to the standard EnGMF in low-ensemble regimes.
Fonte: arXiv cs.LG
Evaluation/Benchmarks • Score 85
Token Arena: A Continuous Benchmark Unifying Energy and Cognition in AI Inference
arXiv:2605.00300v1 Announce Type: new
Abstract: Public inference benchmarks compare AI systems at the model and provider level, but the unit at which deployment decisions are actually made is the endpoint: the (provider, model, stock-keeping-unit) tuple at which a specific quantization, decoding strategy, region, and serving stack is exposed. We introduce TokenArena, a continuous benchmark that measures inference at endpoint granularity along five core axes (output speed, time to first token, workload-blended price, effective context, and quality on the live endpoint) and synthesizes them, together with a modeled energy estimate, into three headline composites: joules per correct answer, dollars per correct answer, and endpoint fidelity (output-distribution similarity to a first-party reference). The framework's novelty is empirical and methodological. Across 78 endpoints serving 12 model families, the same model on different endpoints differs in mean accuracy by up to 12.5 points on math and code, in fingerprint similarity to first party by up to 12 points, in tail latency by an order of magnitude, and in modeled joules per correct answer by a factor of 6.2. We further show that workload-aware blended pricing reorders the leaderboard substantially: 7 of 10 top-ranked endpoints under the chat preset (3:1 input:output) fall out of the top 10 under the retrieval-augmented preset (20:1), and the reasoning preset (1:5) elevates frontier closed models that the chat preset penalizes on price. We release the framework, schema, probe and eval harness, and a v1.0 leaderboard snapshot under CC BY 4.0. TokenArena is a methodology, not a single ranking; we publish full provenance and limitations and welcome external replication.
Fonte: arXiv cs.AI
RL • Score 85
Middle-mile logistics through the lens of goal-conditioned reinforcement learning
arXiv:2605.02461v1 Announce Type: new
Abstract: Middle-mile logistics describes the problem of routing parcels through a network of hubs linked by trucks with finite capacity. We rephrase this as a multi-object goal-conditioned MDP. Our method combines graph neural networks with model-free RL, extracting small feature graphs from the environment state.
Fonte: arXiv stat.ML
Theory/Optimization • Score 85
Spectral Graph Sparsification Preserves Representation Geometry in Graph Neural Networks
arXiv:2605.01136v1 Announce Type: new
Abstract: Spectral graph sparsification is a classical tool for reducing graph complexity while preserving Laplacian quadratic forms. In graph neural networks (GNNs), sparsification is often used to accelerate computation while maintaining predictive performance. In this work, we study a complementary representation-level question: does sparsification preserve the geometry of learned embeddings?
For polynomial-filter GNNs, we prove that any $\epsilon$-spectral sparsifier induces $O(\epsilon)$ perturbations in polynomial graph filters, multilayer hidden representations, and their Gram matrices. These guarantees imply stability of squared pairwise distances, class means, and covariance structure in embedding space. We further establish finite-time training stability: under smoothness and boundedness assumptions, gradient descent on dense and sparsified graphs produces weight trajectories whose separation grows at most proportionally to the sparsification distortion.
Empirically, effective-resistance sparsification validates the predicted perturbation chain on synthetic graphs and preserves hidden representation geometry on real datasets. In our experiments, the gram matrix and training dynamics show low divergence even under substantial sparsification, consistent with the predicted stability under spectral sparsification. Hidden Gram preservation strongly predicts neighborhood preservation and class-centroid stability across FashionMNIST, Cora, and Paul15. Together, these results show that spectral sparsification preserves not only graph operators, but also the representation geometry that supports downstream use of GNN embeddings for interpretability.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Bayesian Optimization in Linear Time
arXiv:2605.00237v1 Announce Type: new
Abstract: Bayesian optimization is a sequential method for minimizing objective functions that are expensive to evaluate and about which few assumptions can be made. By using all gathered data to train a Gaussian process model for the function and adaptively employing a mixture of global exploration and local exploitation, this method has been used for optimization in many fields including machine learning, automotive engineering and reinforcement learning. However, the standard method suffers from two problems: 1) with cubic computational complexity in the training-set size it eventually becomes computationally infeasible to train the model, and 2) globally modeling the objective function is not necessarily optimal given the local nature of minimization. Using flexible and recursive binary partitioning of the search space, we adapt both the modeling and acquisitive aspects of standard Bayesian optimization to work harmoniously with the partitioning scheme, thereby ameliorating both standard shortcomings. We compare our method against a commonly used Bayesian optimization library on seven challenging test functions, ranging in dimensionality from $6$ to $124$, and show that our method achieves superior optimization performance in all tests. In addition our method has linear computational complexity.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Borrowed Geometry: Computational Reuse of Frozen Text-Pretrained Transformer Weights Across Modalities
arXiv:2605.00333v1 Announce Type: new
Abstract: Frozen Gemma 4 31B weights pretrained exclusively on text tokens, unmodified, transfer across modality boundaries through a thin trainable interface. (1) OGBench scene-play-singletask-task1-v0: $+4.33$pt over published GCIQL at $n=3$ with std 0.74 -- a published-SOTA win on a robotic manipulation task the substrate has never seen. (2) D4RL Walker2d-medium-v2: Decision-Transformer parity ($76.2 \pm 0.8$, $n=3$) at $0.43\times$ DT's trainable count, with the frozen substrate compressing to a 5L slice ($+1.66$pt over the 6L baseline at $n=3$). (3) Associative recall as the cleanest pretraining-load-bearing case: the frozen slice + a 113K-parameter linear interface reaches L30 best-checkpoint per-bit error 0.0505 ($n=2$); a 6.36M-parameter from-scratch trained transformer at matched capacity ($1/\sqrt{d_k}$ scaling, two seeds, LR sweep) cannot solve the task at all under the protocol (best L30 = 0.4395), an $8.7\times$ advantage. Architecture-alone falsifications: a frozen random transformer with correct $1/\sqrt{d_k}$ scaling stays at random-chance loss for 50k steps; a random-init Gemma slice fails OGBench cube-double-play-task1 entirely (0.89% across $n=3$ where pretrained reaches 60%). A dual-measurement protocol -- text-activation probing on 95 English sentences plus task-ablation on a non-language target -- names individual heads independently identifiable on both protocols: head L26.28 scores $3.7\times$ the slice mean for English token-copying and is the #2 most-critical head for binary copy ablation ($\Delta$ L30 $= +0.221$); three further heads (L27.28, L27.2, L27.3) classify by the same protocol. The mechanism is single-model and the cross-modality results are single-task within their respective benchmarks; cross-model replication is structurally constrained because Gemma 4 31B is the only model on the small-scale Pareto frontier as of April 2026.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Decentralized Proximal Stochastic Gradient Langevin Dynamics
arXiv:2605.00723v1 Announce Type: new
Abstract: We propose Decentralized Proximal Stochastic Gradient Langevin Dynamics (DE-PSGLD), a decentralized Markov chain Monte Carlo (MCMC) algorithm for sampling from a log-concave probability distribution constrained to a convex domain. Constraints are enforced through a shared proximal regularization based on the Moreau-Yosida envelope, enabling unconstrained updates while preserving consistency with the target constrained posterior. We establish non-asymptotic convergence guarantees in the 2-Wasserstein distance for both individual agent iterates and their network averages. Our analysis shows that DE-PSGLD converges to a regularized Gibbs distribution and quantifies the bias introduced by the proximal approximation. We evaluate DE-PSGLD for different sampling problems on synthetic and real datasets. As the first decentralized approach for constrained domains, our algorithm exhibits fast posterior concentration and high predictive accuracy.
Fonte: arXiv stat.ML
Theory/Optimization • Score 85
A Dirac-Frenkel-Onsager principle: Instantaneous residual minimization with gauge momentum for nonlinear parametrizations of PDE solutions
arXiv:2605.00284v1 Announce Type: new
Abstract: Dirac-Frenkel instantaneous residual minimization evolves nonlinear parametrizations of PDE solutions in time, but ill-conditioning can render the parameter dynamics non-unique.
We interpret this non-uniqueness as a gauge freedom: nullspace directions that leave the
time derivative unchanged can be used to select better-conditioned parameter velocities.
Building on Onsager's minimum-dissipation principle, we introduce a history variable -- interpretable as momentum -- and inject it only along the nullspace directions.
The resulting Dirac-Frenkel-Onsager dynamics preserve instantaneous residual minimization, in contrast to standard regularization that can introduce bias, while promoting temporally smooth parameter evolutions.
Examples demonstrate that the approach leads to increased robustness in singular and near-singular regimes.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Conformalized Quantum DeepONet Ensembles for Scalable Operator Learning with Distribution-Free Uncertainty
arXiv:2605.00330v1 Announce Type: new
Abstract: Operator learning enables fast surrogate modeling of high-dimensional dynamical systems, but existing approaches face two fundamental limitations: quadratic inference complexity and unreliable uncertainty quantification in safety-critical settings. We propose Conformalized Quantum DeepONet Ensembles, a framework that addresses both challenges simultaneously. By leveraging Quantum Orthogonal Neural Networks (QOrthoNNs), we reduce operator inference complexity from O(n^2) to O(n), enabling scalable evaluation over fine discretizations. To provide rigorous uncertainty quantification, we combine ensemble-based epistemic modeling with adaptive conformal prediction, yielding distribution-free coverage guarantees. A key challenge in ensembling is that naive parallelism scales hardware resources linearly with the number of models. We resolve this by using Superposed Parameterized Quantum Circuits (SPQCs), which compress multiple ensemble members into a single circuit and enable simultaneous multi-model execution. Experiments on synthetic partial differential equations and real-world power system dynamics demonstrate that our approach achieves accurate predictions while maintaining calibrated uncertainty under realistic quantum noise. These results establish a practical pathway toward scalable, uncertainty-aware operator learning in quantum machine learning.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
AGoQ: Activation and Gradient Quantization for Memory-Efficient Distributed Training of LLMs
arXiv:2605.00539v1 Announce Type: new
Abstract: Quantization is a key method for reducing the GPU memory requirement of training large language models (LLMs). Yet, current approaches are ineffective for 4-bit activations and 8-bit gradients, which would easily cause slow convergence or accuracy loss. To address this, we introduce AGoQ, incorporating two new techniques: 1) a layer-aware activation quantization algorithm that allocates appropriate bit-widths for activations of various layers based on their types and pipeline stages to achieve near 4-bit activation storage, and 2) a gradient quantization algorithm that reduces memory usage and shortens communication time by employing 8-bit gradient storage and precision-preserving 8-bit All-Reduce communication. We conduct extensive experiments using different sizes of LLMs on two GPU clusters (up to 64 GPUs), and the experimental results show that our AGoQ reduces the memory by up to 52\% and achieves up to 1.34$\times$ improvement of training speed compared to state-of-the-art training systems Megatron-LM (w/ or w/o ZeRO), COAT and DeepSpeed with 8B to 32B LLaMA models, while achieving convergence loss on pretraining and comparable accuracy on downstream tasks with LLaMA architectures.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
SPLICE: Latent Diffusion over JEPA Embeddings for Conformal Time-Series Inpainting
arXiv:2605.00126v1 Announce Type: new
Abstract: Generative models for time-series imputation achieve strong reconstruction accuracy, yet provide no finite-sample reliability guarantees, a critical limitation in power systems where imputed values inform dispatch and planning. We introduce SPLICE (Self-supervised Predictive Latent Inpainting with Conformal Envelopes), a modular framework coupling latent generative imputation with distribution-free, online-adaptive prediction intervals. A JEPA encoder maps daily load segments into a 64-dimensional latent space; a conditional latent bridge with four sampling modes generates candidate gap trajectories; an hourly-conditioned decoder maps back to signal space; and Adaptive Conformal Inference (ACI) wraps the output with coverage-guaranteed prediction bands. The flow-matching variant achieves comparable quality to DDIM in 5--10 ODE steps (5-10x speedup). On thirteen load datasets (nine proprietary, three UCI Electricity, ETTh1), SPLICE achieves the lowest mean Load-only MSE (0.056), winning 9/12 non-degenerate datasets at 91-day gaps and 18/32 across all gap lengths vs. five established baselines, and produces the best CRPS (0.161, -18.3% vs. the strongest competitor). ACI delivers 93--95% empirical coverage, correcting under-coverage failures of up to 7.5 pp observed with static conformal prediction. A pooled JEPA encoder trained on nine feeds transfers to four unseen domains, matching or exceeding per-dataset oracles with only a quick bridge fine-tuning.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
High-Probability Convergence in Decentralized Stochastic Optimization with Gradient Tracking
arXiv:2605.00281v1 Announce Type: new
Abstract: We study high-probability (HP) convergence guarantees in decentralized stochastic optimization, where multiple agents collaborate to jointly train a model over a network. Existing HP results in decentralized settings almost exclusively focus on the Decentralized Stochastic Gradient Descent ($\mathtt{DSGD}$) algorithm, which requires strong assumptions, such as bounded data heterogeneity, or strong convexity of each agent's cost. This is contrary to the mean-squared error (MSE) results, where methods incorporating bias-correction techniques are known to converge under relaxed assumptions and achieve better practical performance. In this paper we provide the first step toward bridging the gap, by studying HP convergence of $\mathtt{DSGD}$ incorporating the gradient tracking technique, in the presence of noise satisfying a relaxed sub-Gaussian condition. We show that the resulting method, dubbed $\mathtt{GT-DSGD}$, achieves order-optimal HP convergence rates for both non-convex and Polyak-\L{}ojasiewicz costs, of order $\mathcal{O}\Big(\frac{\log(1/\delta)}{\sqrt{nT}}\Big)$ and $\mathcal{O}\Big(\frac{\log(1/\delta)}{nT}\Big)$, respectively, where $n$ is the number of agents, $T$ is the time horizon and $\delta \in (0,1)$ is the confidence parameter. Our results establish that $\mathtt{GT-DSGD}$ converges in the HP sense under the same conditions on the cost as in the MSE sense, while achieving comparable transient times. To the best of our knowledge, these are the first HP guarantees for decentralized optimization methods incorporating bias-correction. Numerical experiments on real and synthetic data verify our theoretical findings, underlining the superior performance of $\mathtt{GT-DSGD}$ and highlighting that the benefits of incorporating bias-correction are also maintained in the HP sense.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
MemRouter: Memory-as-Embedding Routing for Long-Term Conversational Agents
arXiv:2605.00356v1 Announce Type: new
Abstract: Long-term conversational agents must decide which turns to store in external memory, yet recent systems rely on autoregressive LLM generation at every turn to make that decision. We present MemRouter, a write-side memory router that decouples memory admission from the downstream answer backbone and replaces per-turn memory-management decoding with an embedding-based routing policy. MemRouter encodes each turn together with recent context, projects the resulting embeddings through a frozen LLM backbone, and predicts whether the turn should be stored using lightweight classification heads while training only 12M parameters. Under a controlled matched-harness comparison on LoCoMo, where the retrieval pipeline, answer prompts, and QA backbone (Qwen2.5-7B) are held identical, MemRouter outperforms an LLM-based memory manager on every question category (overall F1 52.0 vs 45.6, non-overlapping 95% CIs) while reducing memory-management p50 latency from 970ms to 58ms. Descriptive factorial averaging further shows that learned admission improves mean F1 by +10.3 over random storage, category-specific prompting adds +5.2 over a generic prompt, and retrieval contributes +0.7. These results suggest that write-side memory admission can be learned by a small supervised router, while answer generation remains a separate downstream component in long-horizon conversational QA.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
Scaling Video Understanding via Compact Latent Multi-Agent Collaboration
arXiv:2605.00444v1 Announce Type: new
Abstract: Multi-modal large language models (MLLMs) advance vision language understanding but face inherent limitations in long-video tasks due to bounded perception context budgets. Existing agentic methods mitigate this via rule-based preprocessing, yet often suffer from information loss, high cost, and reliance on textual intermediates. We propose MACF, an end-to-end Multi-Agent Collaboration Framework that decouples per-agent perception budgets from global video complexity, enabling scalable video understanding while preserving visual fidelity. MACF partitions videos into segments for locally budgeted agents and enables holistic reasoning via an agent-native latent communication protocol. Each agent encodes partial observations into compact, task-sufficient tokens in a shared embedding space, allowing efficient and information-preserving collaboration by a central coordinator. We introduce a curriculum training strategy that progressively enforces semantic alignment, evidence summarization, and cross-agent coordination. Extensive experiments on diverse video understanding benchmarks show that MACF consistently outperforms state-of-the-art MLLMs and multi-agent systems under identical budget constraints, demonstrating the effectiveness of our latent collaboration for scalable video understanding.
Fonte: arXiv cs.CV
Theory/Optimization • Score 85
NLPOpt-Net: A Learning Method for Nonlinear Optimization with Feasibility Guarantees
arXiv:2605.00260v1 Announce Type: new
Abstract: Nonlinear Parametric Optimization Network (NLPOpt-Net) is an unsupervised learning architecture to solve constrained nonlinear programs (NLP). Given the structure of an NLP, it learns the parametric solution maps with guaranteed constraint satisfaction. The architecture consists of a backbone neural network (NN) followed by a multilayer ($k$-layered) projection. While the NN drives toward optimality through a loss function consisting of a modified Lagrangian augmented with a consistency loss, the projection ensures feasibility by projecting the NN predictions in the original constraint manifold. Instead of typical distance minimization, our projection exploits local quadratic approximations of the original NLP. Under certain conditions (such as convexity), the projection has a descent property, which improves the NN predictions further. NLPOpt-Net deploys an inversion-free, modified Chambolle-Pock algorithm to solve the constrained quadratic projections during the forward pass and uses the implicit function theorem for efficient backpropagation. The fixed structure of the projection further allows decoupling of the NN and the projection once the training is complete. NLPOpt-Net solves large-scale convex QP, QCQP, NLP, and nonconvex problems with near zero optimality gap and constraint violations reduced to machine precision. Additionally, it provides near accurate prediction of the active sets and corresponding dual variables, thereby enabling a scalable approach for multiparametric programming. Compiling the projection in C provides order of magnitude improvement in inference time compared to JAX. We provide the codes and NLPOpt-Net as a ready to use package that includes GPU support.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Federated Weather Modeling on Sensor Data
arXiv:2605.00322v1 Announce Type: new
Abstract: Federated weather modeling on sensor data is a distributed system underpinned by federated learning, enabling multiple sensor data sources, including ground weather stations, satellites and IoT devices, to collaboratively train deep learning models without sharing raw data. This method safeguards data privacy and security while leverages diverse, geographically distributed datasets to improve the accuracy and robustness of global/regional weather modeling tasks such as forecasting and anomaly detection.
Fonte: arXiv cs.LG
RecSys • Score 85
Smart Ensemble Learning Framework for Predicting Groundwater Heavy Metal Pollution
arXiv:2605.00056v1 Announce Type: new
Abstract: Groundwater in the Densu Basin is increasingly threatened by heavy metal contamination, but conventional methods fail to capture the statistical complexity and spatial heterogeneity of pollution indicators. A key challenge is modelling the Heavy Metal Pollution Index (HPI), which is typically skewed and affected by correlated contaminants, leading to biased predictions without transformation. This study develops a predictive framework integrating response transformations with nested cross-validated ensemble machine learning. Three transformations (raw, log, and Gaussian copula) were applied to HPI and evaluated across six learners: support vector regression (SVM), $k$-nearest neighbours (k-NN), CART, Elastic Net, kernel ridge regression, and a stacked Lasso ensemble. Raw-scale models produced deceptively high fits (Elastic Net and stacked ensemble $R^2 \approx 1.0$), suggesting over-optimism. The log transformation stabilised variance (SVM: $R^2 = 0.93$, RMSE $= 0.18$; k-NN: $R^2 = 0.92$, RMSE $= 0.20$). The Gaussian copula gave the most reliable results: stacked ensemble $R^2 = 0.96$ (RMSE $= 0.19$), with other learners maintaining high accuracy. Copula-based models improved residuals and produced spatially plausible maps. DBSCAN clustering revealed Fe and Mn as primary HPI contributors, consistent with regional hydrogeochemistry. Limitations include reliance on random (not spatial) cross-validation and basin-specific scope. Future work should explore spatial validation and other geological settings. Overall, distribution-aware ensembles with clustering diagnostics offer robust, interpretable assessments of groundwater contamination.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
FedACT: Concurrent Federated Intelligence across Heterogeneous Data Sources
arXiv:2605.00011v1 Announce Type: new
Abstract: Federated Learning (FL) enables collaborative intelligence across decentralized data source devices in a privacy-preserving way. While substantial research attention has been drawn to optimizing the learning process for an individual task, real-world applications increasingly require multiple machine learning tasks simultaneously training their models across a shared pool of devices. Naively applying single-FL optimization techniques in multi-FL systems results in suboptimal system performance, particularly due to device heterogeneity and resource inefficiency. To address such a critical open challenge, we introduce {\em FedACT}, a novel resource heterogeneity-aware device scheduling approach designed to efficiently schedule heterogeneous devices across multiple concurrent FL jobs, with the goal of minimizing their average job completion time (JCT). {\em FedACT} dynamically assigns devices to FL jobs based on an alignment scoring mechanism that evaluates the compatibility between available resources of devices and resource demands of jobs. Additionally, it incorporates participation fairness to ensure balanced contributions from devices across jobs, further enhancing the accuracy levels of learned global models. An optimal scheduling plan is formulated in {\em FedACT} by prioritizing devices with higher alignment scores, while ensuring fair participation across jobs. To evaluate the effectiveness of the proposed scheduling algorithm, we carried out comprehensive experiments using diverse FL jobs and benchmark datasets. Experimental results demonstrate that {\em FedACT} reduces the average JCT by up to 8.3\(\times\) and improves model accuracy by up to 44.5\%, compared to the state-of-the-art baselines.
Fonte: arXiv cs.LG
Vision • Score 85
BOLT: Online Lightweight Adaptation for Preparation-Free Heterogeneous Cooperative Perception
arXiv:2605.00405v1 Announce Type: new
Abstract: Most existing heterogeneous cooperative perception methods depend on prior preparation like offline joint training or tailored collaborator-model adaptation. Such preprocessing is, however, generally impractical in real scenarios, as agents are usually independently trained by different developers and meet occasionally online. This work investigates \emph{preparation-free heterogeneous cooperative perception}, where agents use independently trained single-agent detectors without any pre-deployment coordination. We find direct cross-agent fusion under this setting greatly underperforms ego-only perception. We present BOLT, a lightweight plug-and-play module that adapts neighboring features online via ego-as-teacher distillation, requiring only ego predictions without ground-truth labels. BOLT leverages high-confidence ego perception features to guide cross-agent feature-domain alignment, while enabling neighbors to contribute features in the ego's low-confidence regions. With only 0.9M trainable parameters, BOLT improves AP@50 by up to 32.3 points over vanilla unadapted fusion in the preparation-free setting. It consistently outperforms ego-only results on DAIR-V2X and OPV2V, across different encoder pairs and fusion strategies. Code: https://github.com/sidiangongyuan/BOLT.
Fonte: arXiv cs.CV
Applications • Score 85
Smart Profit-Aware Crop Advisory System: Kisan AI
arXiv:2605.00133v1 Announce Type: new
Abstract: Modern crop advisory systems exhibit a critical limitation termed \textit{economic blindness}. These systems primarily optimize for biological yield, often overlooking market price, which can lead farmers toward agronomically sound yet financially unviable decisions. In this paper, we develop Kisan AI, a smart profit-aware crop advisory system that resolves the above-mentioned limitation through a research-driven, full-stack application. We train the Random Forest(RF) classifier model on a nine-feature benchmark dataset, the standard seven agronomic attributes augmented with a \textit{market\_price} variable, and evaluated against eight baseline models, considering the evaluation matrices, such as, accuracy, precision, recall, F1-score, and Log Loss. The RF model achieves the highest accuracy of 99.3\% and the lowest Log Loss, confirming that the inclusion of market price as a predictive feature is both valid and impactful. We then implement the RF model within a multilingual progressive Web App alongside a Facebook Prophet six-month price forecasting engine and a MobileNetV2 disease detection module. A nine-language AI chatbot powered by the Anthropic Claude API unifies all modules into a single, mobile-installable platform accessible to farmers across India.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
AirFM-DDA: Air-Interface Foundation Model in the Delay-Doppler-Angle Domain for AI-Native 6G
arXiv:2605.00020v1 Announce Type: new
Abstract: The success of large foundation models is catalyzing a new paradigm for AI-native 6G network design: wireless foundation models for physical layer design. However, existing models often operate on channel state information (CSI) in the space-time-frequency (STF) domain, where distinct multipath components are inherently superimposed and structurally entangled. This hinders the learning of universal channel representation. Meanwhile, their reliance on global attention mechanisms incurs prohibitive computational overhead. In this paper, we propose AirFM-DDA, an Air-interface Foundation Model operating in the Delay-Doppler-Angle (DDA) domain for physicallayer tasks. Specifically, AirFM-DDA reparameterizes CSI from the STF domain into the DDA domain to explicitly resolve multipath components along physically meaningful axes. It employs a window-based attention module augmented with framestructure-aware positional encoding (FS-PE). This window-based attention aligns with locally clustered multipath dependencies while avoiding quadratic-complexity global attention, and FS-PE injects frame-structure priors into network. Extensive experiments demonstrate that AirFM-DDA achieves superior zero-shot generalization across unseen scenarios and datasets, consistently outperforming the baselines on channel prediction and estimation tasks. Compared to the global attention, its window-based attention reduces training and inference costs by nearly an order of magnitude. Moreover, AirFM-DDA maintains robustness under high mobility, large delay spreads, severe noise, and extreme aliasing conditions.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
VkSplat: High-Performance 3DGS Training in Vulkan Compute
arXiv:2605.00219v1 Announce Type: new
Abstract: We present VkSplat, a high-performance, cross-vendor 3D Gaussian Splatting (3DGS) training pipeline implemented fully in Vulkan compute, addressing performance and compatibility limitation of existing training pipelines. With various optimizations, we achieve $3.3\times$ speed and $33\%$ VRAM reduction over CUDA+PyTorch baseline, maintaining quality, and demonstrating compatibility across GPU vendors. To the best of our knowledge, this is the first fully-Vulkan-based 3DGS training pipeline that achieves state-of-the-art performance. Code: \href{https://github.com/harry7557558/vksplat}{https://github.com/harry7557558/vksplat}
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
Learning from Compressed CT: Feature Attention Style Transfer and Structured Factorized Projections for Resource-Efficient Medical Image Analysis
arXiv:2605.00448v1 Announce Type: new
Abstract: The deployment of artificial intelligence in medical imaging is hindered by high computational complexity and resource-intensive processing of volumetric data. Although chest computed tomography (CT) volumes offer richer diagnostic information than projection radiography, their use in AI-based diagnosis remains limited due to the computational burden of processing uncompressed volumetric images (typically stored in NIfTI or DICOM format). Addressing the growing need for low-resource deployment and efficient electronic data transfer, we investigate the utilization of JPEG-compressed chest CT volumes for thoracic abnormality detection. We propose Feature Attention Style Transfer (FAST), a novel distillation framework that transfers both activation patterns and structural relationships from high-fidelity CT representations to a spatiotemporal visual encoder operating on compressed inputs. By combining Gram-matrix-based attention style preservation with dual-attention feature alignment, FAST enables robust feature extraction from degraded volumes. Furthermore, we introduce Structured Factorized Projection (SFP), leveraging Block Tensor Train decomposition as a parameter-efficient alternative to dense projection layers, reducing projection-head parameters by almost half. Our contrastive learning pipeline, CT-Lite, integrates these components with a SigLIP-based multimodal alignment objective. Experiments on CT-RATE, NIDCH, and Rad-ChestCT demonstrate that CT-Lite achieves AUROC within 5-7\% of the uncompressed-input baseline across all three datasets, despite operating on compressed inputs with significantly fewer parameters, paving the way for AI-based clinical evaluation under resource constraints.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
A11y-Compressor: A Framework for Enhancing the Efficiency of GUI Agent Observations through Visual Context Reconstruction and Redundancy Reduction
arXiv:2605.00551v1 Announce Type: new
Abstract: AI agents that interact with graphical user interfaces (GUIs) require effective observation representations for reliable grounding. The accessibility tree is a commonly used text-based format that encodes UI element attributes, but it suffers from redundancy and lacks structural information such as spatial relationships among elements. We propose A11y-Compressor, a framework that transforms linearized accessibility trees into compact and structured representations. Our implementation, Compressed-a11y, applies a lightweight and structured transformation pipeline with modal detection, redundancy reduction, and semantic structuring. Experiments on the OSWorld benchmark show that Compressed-a11y reduces input tokens to 22% of the original while improving task success rates by 5.1 percentage points on average.
Fonte: arXiv cs.CL
Theory/Optimization • Score 85
Towards Robust and Scalable Density-based Clustering via Graph Propagation
arXiv:2605.00390v1 Announce Type: new
Abstract: We present \textit{CluProp}, a novel framework that reimagines varied-density clustering in high-dimensional spaces as a label propagation process over neighborhood graphs. Our approach formally bridges the gap between density-based clustering and graph connectivity, leveraging efficient propagation mechanisms from network science to mitigate the parameter sensitivity inherent in traditional density-based methods. Specifically, we introduce a deterministic density-based propagation strategy to ensure scalable neighborhood identification. The framework is agnostic to the choice of distance metric and exhibits superior performance on large-scale data, processing millions of points in minutes while consistently outperforming existing baselines in accuracy.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Real-Time Frame- and Event-based Object Detection with Spiking Neural Networks on Edge Neuromorphic Hardware: Design, Deployment and Benchmark
arXiv:2605.00146v1 Announce Type: new
Abstract: Real-time object detection on energy-constrained platforms is critical for applications such as UAV-based inspection, autonomous navigation, and mobile robotics. Spiking neural networks (SNNs) on neuromorphic hardware are believed to be significantly more energy-efficient than conventional artificial neural networks (ANNs). In this work, we present a comprehensive methodology for designing general SNN detection architectures targeting neuromorphic platforms, along with the engineering adaptations required to deploy them on the state-of-the-art Neuromorphic processor, Intel Loihi 2. We benchmark SNN-based object detection on Loihi 2 using both frame-based and event-based datasets, comparing performance with ANN-based detection on the NVIDIA Jetson Orin Nano, NVIDIA Jetson Nano B01, and the Apple M2 CPU. Our results show that SNNs on Loihi 2 can perform real-time detection while achieving the lowest per-inference dynamic energy among all platforms. Also, Loihi 2 outperforms the other platforms in terms of power consumption, though ANNs on Jetson Orin Nano achieve higher inference rates. Furthermore, our ANN-to-SNN distillation-aware training enables SNNs to recover 87-100% of the detection accuracy of their ANN counterparts while maintaining lower inference latency; without distillation, SNNs exhibit an 11-27% accuracy drop. These results highlight the potential of neuromorphic systems for energy-efficient, real-time object detection at the edge.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
CRADIPOR: Crash Dispersion Predictor
arXiv:2605.00070v1 Announce Type: new
Abstract: We present CRADIPOR, a numerical dispersion prediction tool for automotive crash simulations. Finite Element (FE) crash models are widely used throughout vehicle development, but their predictions are not strictly repeatable because of parallel computation and model complexity. As a result, performance criteria evaluated during post-processing may exhibit significant numerical dispersion, which complicates engineering decision-making. Although dispersion can be estimated by repeating the same simulation, this approach is generally impractical because of its high computational cost.
This work therefore investigates a prediction tool that can be applied during routine crash-simulation post-processing without repeating the computation. The proposed approach relies on a Rank Reduction Autoencoder (RRAE) combined with supervised classification in order to identify regions sensitive to numerical dispersion. The comparative analysis suggests that the RRAE-based framework is more effective than the Random Forest baseline on the studied dataset. Among the tested signal representations, wavelet-based and slope-based inputs appear to be the most promising, with slope variations providing the best classification performance. These results support the use of structured latent representations for improving numerical-dispersion detection in automotive crash post-processing.
Fonte: arXiv cs.LG
Applications • Score 85
Agentic AI for Substance Use Education: Integrating Regulatory and Scientific Knowledge Sources
arXiv:2605.00383v1 Announce Type: new
Abstract: The delivery of traditional substance education has remained problematic due to challenges in scalability, personalization, and the currency of information in a rapidly evolving substance use landscape. While artificial intelligence (AI) offers a promising frontier for enhancing educational delivery, its application in providing real-time, authoritative substance use education remains largely underexplored. We built an agentic-based AI web application that combined Drug Enforcement Administration records with peer-reviewed literature in real-time to provide transparent context-sensitive substance use education. The system uses retrieval-augmented generation with a carefully filtered corpus of 102 documents and dynamic PubMed queries. Document storage was semantically chunked and placed in a vector representation in order to be easily retrieved. We conducted an expert evaluation study in which a panel of five subject matter experts generated 30 domain-specific questions, and two independent raters assessed 90 system interactions (30 primary questions plus two contextual follow-ups each) using a five-point Likert scale across four criteria: factual accuracy, citation quality, contextual coherence, and regulatory appropriateness. Mean ratings ranged from 4.18 to 4.35 across the four criteria (overall category range: 4.05-4.52), with substantial inter-rater agreement (Cohen's kappa = 0.78). These findings suggest that agentic AI architectures integrating authoritative regulatory sources with real-time scientific literature represent a promising direction for scalable, accurate, and verifiable health education delivery, warranting further evaluation through longitudinal user studies.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
RadLite: Multi-Task LoRA Fine-Tuning of Small Language Models for CPU-Deployable Radiology AI
arXiv:2605.00421v1 Announce Type: new
Abstract: Large language models (LLMs) show promise in radiology but their deployment is limited by computational requirements that preclude use in resource-constrained clinical environments. We investigate whether small language models (SLMs) of 3-4 billion parameters can achieve strong multi-task radiology performance through LoRA fine-tuning, enabling deployment on consumer-grade CPUs. We train Qwen2.5-3B-Instruct and Qwen3-4B on 162K samples spanning 9 radiology tasks - RADS classification across 10 systems, impression generation, temporal comparison, radiology NLI, NER, abnormality detection, N/M staging, and radiology Q&A - compiled from 12 public datasets. Both models are evaluated on up to 500 held-out test samples per task with standardized metrics. Our key findings are: (1) LoRA fine-tuning dramatically improves performance over zero-shot baselines (RADS accuracy +53%, NLI +60%, N-staging +89%); (2) the two models exhibit complementary strengths - Qwen2.5 excels at structured generation tasks while Qwen3 dominates extractive tasks; (3) a task-outed oracle ensemble combining both models achieves the best performance across all tasks; (4) few-shot prompting with fine-tuned models hurts performance, demonstrating that LoRA adaptation is more effective than in-context learning for specialized domains; and (5) models can be quantized to GGUF format (~1.8-2.4GB) for CPU deployment at 4-8 tokens/second on consumer hardware. Our work demonstrates that small, efficiently fine-tuned models - which we collectively call RadLite - can serve as practical multi-task radiology AI assistants deployable entirely on consumer hardware without GPU requirements.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
Federated Distillation for Whole Slide Image via Gaussian-Mixture Feature Alignment and Curriculum Integration
arXiv:2605.00578v1 Announce Type: new
Abstract: Federated learning (FL) offers a promising framework for collaborative digital pathology by enabling model training across institutions. However, real-world deployments face heterogeneity arising from diverse multiple instance learning (MIL) architectures and heterogeneous feature extractors across institutions. We propose FedHD, a novel FL framework that performs local Gaussian-mixture feature alignment tailored for WSI analysis. Instead of exchanging model parameters, each client independently distills semantically rich synthetic feature representations aligned with the distribution of real WSIs. To preserve diagnostic diversity, FedHD adopts a one-to-one distillation strategy, generating a synthetic counterpart for each real slide to avoid over-compression. During federation, a curriculum-based integration strategy progressively incorporates cross-site synthetic features into local training once performance plateaus. Furthermore, an optional interpretation module reconstructs pseudo-patches from synthetic embeddings, enhancing transparency. FedHD is architecture-agnostic, privacy-preserving, and supports personalized yet collaborative training across diverse institutions. Experiments on TCGA-IDH, CAMELYON16, and CAMELYON17 show that FedHD consistently outperforms state-of-the-art federated and distillation baselines.
Fonte: arXiv cs.CV
Vision • Score 85
Leveraging Vision-Language Models as Weak Annotators in Active Learning
arXiv:2605.00480v1 Announce Type: new
Abstract: Active learning aims to reduce annotation cost by selectively querying informative samples for supervision under a limited labeling budget. In this work, we investigate how vision-language models (VLMs) can be leveraged to further reduce the reliance on costly human annotation within the active learning paradigm. To this end, we find that the reliability of VLMs varies significantly with label granularity in fine-grained recognition tasks: they perform poorly on fine-grained labels but can provide accurate coarse-grained labels. Leveraging this property, we propose an active learning framework that combines fine-grained human annotations with coarse-grained VLM-generated weak labels through instance-wise label assignment. We further model the systematic noise in VLM-generated labels using a small set of trusted full labels. Experiments on CUB200 and FGVC-Aircraft show that the proposed framework consistently outperforms existing active learning methods under the same annotation budget.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Agent Capsules: Quality-Gated Granularity Control for Multi-Agent LLM Pipelines
arXiv:2605.00410v1 Announce Type: new
Abstract: A multi-agent pipeline with N agents typically issues N LLM calls per run. Merging agents into fewer calls (compound execution) promises token savings, but naively merged calls silently degrade quality through tool loss and prompt compression. We present Agent Capsules, an adaptive execution runtime that treats multi-agent pipeline execution as an optimization problem with empirical quality constraints. The runtime instruments coordination overhead per group, scores composition opportunity, selects among three compound execution strategies, and gates every mode switch on rolling-mean output quality. A controlled negative result confirms that injecting more context into a merged call worsens compression rather than relieving it, so the framework's escalation ladder (standard, then two-phase, then sequential) recovers quality by moving toward per-agent dispatch rather than by rewriting merged prompts. On LLM-judged quality, the controller matches a hand-tuned oracle on every measured (model, group, mode) cell: routing compound whenever the oracle would, and reverting to fine whenever quality would fail the floor, without per-model configuration. Against a hand-crafted LangGraph implementation of a 14-agent competitive intelligence pipeline, Agent Capsules uses 51% fewer fine-mode input tokens and 42% fewer compound-mode input tokens, at +0.020 and +0.017 quality respectively. Against a DSPy implementation of a 5-agent due diligence pipeline, the framework uses 19% fewer tokens than uncompiled DSPy at quality parity, and 68% fewer tokens than MIPROv2 at +0.052 quality. Even before compound mode fires, the runtime delivers efficiency through automatic policy resolution, cache-aligned prompts, and topology-aware context injection, matching both hand-tuned and compile-time baselines without training data or per-pipeline engineering.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
Technical Report: Activation Residual Hessian Quantization (ARHQ) for Low-Bit LLM Quantization
arXiv:2605.00140v1 Announce Type: new
Abstract: We present Activation Residual Hessian Quantization (ARHQ), a post-training weight splitting method designed to mitigate error propagation in low-bit activation-weight quantization. By constructing an input-side residual Hessian from activation quantization residuals (G_x), ARHQ analytically identifies and isolates error-sensitive weight directions into a high-precision low-rank branch. This is achieved via a closed-form truncated SVD on the scaled weight matrix W G^{1/2}_x . Experimental results on Qwen3-4B-Thinking-2507 demonstrate that ARHQ significantly improves layer-wise SNR and preserves downstream reasoning performance on ZebraLogic even under aggressive quantization. The code is available at https://github.com/BeautMoonQ/ARHQ.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Cloud Is Closer Than It Appears: Revisiting the Tradeoffs of Distributed Real-Time Inference
arXiv:2605.00005v1 Announce Type: new
Abstract: The increasing deployment of deep neural networks (DNNs) in cyber-physical systems (CPS) enhances perception fidelity, but imposes substantial computational demands on execution platforms, posing challenges to real-time control deadlines. Traditional distributed CPS architectures typically favor on-device inference to avoid network variability and contention-induced delays on remote platforms. However, this design choice places significant energy and computational demands on the local hardware. In this work, we revisit the assumption that cloud-based inference is intrinsically unsuitable for latency-sensitive control tasks. We demonstrate that, when provisioned with high-throughput compute resources, cloud platforms can effectively amortize network and queueing delays, enabling them to match or surpass on-device performance for real-time decision-making. Specifically, we develop a formal analytical model that characterizes distributed inference latency as a function of the sensing frequency, platform throughput, network delay, and task-specific safety constraints. We instantiate this model in the context of emergency braking for autonomous driving and validate it through extensive simulations using real-time vehicular dynamics. Our empirical results identify concrete conditions under which cloud-based inference adheres to safety margins more reliably than its on-device counterpart. These findings challenge prevailing design strategies and suggest that the cloud is not merely a feasible option, but often the preferred inference location for distributed CPS architectures. In this light, the cloud is not as distant as traditionally perceived; in fact, it is closer than it appears.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Learning Rate Transfer in Normalized Transformers
arXiv:2604.27077v1 Announce Type: new
Abstract: The Normalized Transformer, or nGPT (arXiv:2410.01131) achieves impressive training speedups and does not require weight decay or learning rate warmup. However, despite having hyperparameters that explicitly scale with model size, we observe that nGPT does not exhibit learning rate transfer across model dimension and token horizon. To rectify this, we combine numerical experiments with a principled use of alignment exponents (arXiv:2407.05872) to revisit and modify the $\mu$P approach to hyperparameter transfer (arXiv:2011.14522). The result is a novel nGPT parameterization we call $\nu$GPT. Through extensive empirical validation, we find $\nu$GPT exhibits learning rate transfer across width, depth, and token horizon.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
A Framework for Variational Inference of Lightweight Bayesian Neural Networks with Heteroscedastic Uncertainties
arXiv:2402.14532v2 Announce Type: replace-cross
Abstract: Obtaining heteroscedastic predictive uncertainties from a Bayesian Neural Network (BNN) is vital to many applications. Often, heteroscedastic aleatoric uncertainties are learned as outputs of the BNN in addition to the predictive means, however doing so may necessitate adding more learnable parameters to the network. In this work, we demonstrate that both the heteroscedastic aleatoric and epistemic variance can be embedded into the variances of learned BNN parameters, improving predictive performance for lightweight networks. By complementing this approach with a moment propagation approach to inference, we introduce a relatively simple framework for sampling-free variational inference suitable for lightweight BNNs.
Fonte: arXiv stat.ML
Vision • Score 85
Robust Lightweight Crack Classification for Real-Time UAV Bridge Inspection
arXiv:2604.27617v1 Announce Type: new
Abstract: With the widespread application of Unmanned Aerial Vehicles (UAVs) in bridge structural health monitoring, deep learning-based automatic crack detection has become a major research focus. However, practical UAV inspections still face four key challenges: weak crack features, degraded imaging conditions, severe class imbalance, and limited computational resources for practical UAV inspection workflows. To address these issues, this paper proposes a unified lightweight convolutional neural network framework composed of four synergistic components: a lightweight backbone network, a Convolutional Block Attention Module (CBAM) for channel and spatial enhancement, a directed robust augmentation strategy based on inspection-scene priors, and Focal Loss for hard-sample learning under class imbalance. Experiments on the SDNET2018 bridge deck dataset show that the proposed method achieves an inference speed of 825 FPS with only 11.21M parameters and 1.82G FLOPs. Compared with the baseline model, the complete framework improves the F1-score by 2.51% and recall by 3.95%. In addition, Grad-CAM visualizations indicate that the introduced attention module shifts the model's focus from scattered regions to precise tracking along crack trajectories. Overall, this study achieves a strong balance among accuracy, speed, and robustness, providing a practical solution for ground-station assisted real-time deployment in UAV bridge inspections. The source code is available at: https://github.com/skylynf/AttXNet .
Fonte: arXiv cs.CV
Vision • Score 85
VeraRetouch: A Lightweight Fully Differentiable Framework for Multi-Task Reasoning Photo Retouching
arXiv:2604.27375v1 Announce Type: new
Abstract: Reasoning photo retouching has gained significant traction, requiring models to analyze image defects, give reasoning processes, and execute precise retouching enhancements. However, existing approaches often rely on non-differentiable external software, creating optimization barriers and suffering from high parameter redundancy and limited generalization. To address these challenges, we propose VeraRetouch, a lightweight and fully differentiable framework for multi-task photo retouching. We employ a 0.5B Vision-Language Model (VLM) as the central intelligence to formulate retouching plans based on instructions and scene semantics. Furthermore, we develop a fully differentiable Retouch Renderer that replaces external tools, enabling direct end-to-end pixel-level training through decoupled control latents for lighting, global color, and specific color adjustments. To overcome data scarcity, we introduce AetherRetouch-1M+, the first million-scale dataset for professional retouching, constructed via a new inverse degradation workflow. Furthermore, we propose DAPO-AE, a reinforcement learning post-training strategy that enhances autonomous aesthetic cognition. Extensive experiments demonstrate that VeraRetouch achieves state-of-the-art performance across multiple benchmarks while maintaining a significantly smaller footprint, enabling mobile deployment. Our code and models are publicly available at https://github.com/OpenVeraTeam/VeraRetouch.
Fonte: arXiv cs.CV
Vision • Score 85
CasLayout: Cascaded 3D Layout Diffusion for Indoor Scene Synthesis with Implicit Relation Modeling
arXiv:2604.27361v1 Announce Type: new
Abstract: Synthesizing realistic 3D indoor scenes remains challenging due to data scarcity and the difficulty of simultaneously enforcing global architectural constraints and local semantic consistency. Existing approaches often overlook structural boundaries or rely on fully connected relation graphs that introduce redundant generation errors. Inspired by human design cognition, we present CasLayout, a cascaded diffusion framework that decomposes the joint scene generation task into four conditional sub-stages with explicit physical and semantic roles: (1) predicting furniture quantity and categories, (2) refining object sizes and feature embeddings, (3) modeling spatial relationships in a latent space, and (4) generating Oriented Bounding Boxes (OBBs). This decoupled architecture reduces data requirements and enables flexible integration of Large Language Models (LLMs) and Vision Language Models (VLMs) for zero-shot tasks such as image-to-scene generation. To maintain physical validity within complex floor plans, we explicitly model building elements (e.g., walls, doors, and windows) as conditional constraints. Furthermore, to address the high entropy of dense relation graphs, we introduce a sparse relation graph formulation aligned with human spatial descriptions. By encoding these sparse graphs into a compact latent space using a bidirectional Variational Autoencoder (VAE), the proposed framework provides enhanced relational controllability, allowing generated layouts to better respect functional organization. Experiments demonstrate that CasLayout achieves state-of-the-art performance in fidelity and diversity while enabling improved controllability in practical applications.
Fonte: arXiv cs.CV
Theory/Optimization • Score 85
Compositional Meta-Learning for Mitigating Task Heterogeneity in Physics-Informed Neural Networks
arXiv:2604.26999v1 Announce Type: new
Abstract: Physics-informed neural networks (PINNs) approximate solutions of partial differential equations (PDEs) by embedding physical laws into the loss function. In parameterized PDE families, variations in coefficients or boundary/initial conditions define distinct tasks. This makes training individual PINNs for each task computationally prohibitive, while cross-task transfer can be sensitive to task heterogeneity. While meta-learning can reduce retraining cost, existing methods often rely on a single global initialization and may suffer from negative transfer, particularly under feature-scarce coordinate inputs and limited training-task availability. We propose the Learning-Affinity Adaptive Modular Physics-Informed Neural Network (LAM-PINN), a compositional framework that leverages task-specific learning dynamics. LAM-PINN combines PDE parameters with learning-affinity metrics from brief transfer sessions to construct a task representation and cluster tasks even with coordinate-only inputs. It decomposes the model into cluster-specialized subnetworks and a shared meta network, and learns routing weights to selectively reuse modules instead of relying on a single global initialization. Across three PDE benchmarks, LAM-PINN achieves an average 19.7-fold reduction in mean squared error (MSE) on unseen tasks using only 10% of the training iterations required by conventional PINNs. These results indicate its effectiveness for generalization to unseen configurations within bounded design spaces of parameterized PDE families in resource-constrained engineering settings.
Fonte: arXiv cs.AI
Vision • Score 85
Lightweight Distillation of SAM 3 and DINOv3 for Edge-Deployable Individual-Level Livestock Monitoring and Longitudinal Visual Analytics
arXiv:2604.27128v1 Announce Type: new
Abstract: Foundation-model pipelines for individual-level livestock monitoring -- combining open-vocabulary detection, promptable video segmentation, and self-supervised visual embeddings -- have raised the accuracy ceiling of precision livestock farming (PLF), but their GPU memory budgets exceed the envelope of commodity edge accelerators. To close this gap, the 446M-parameter Perception Encoder (PE-ViT-L+) backbone of SAM 3 is distilled into a 40.66M-parameter multi-scale student through three mechanisms: a Feature Pyramid Network student encoder built on TinyViT-21M-512, a four-term direction-then-scale distillation loss, and backbone-substitution inference with sliding-window session pruning that bounds streaming GPU memory growth. The DINOv3 family includes a pre-distilled ViT-S/16 variant (21.6M parameters) released alongside a 6716M-parameter ViT-7B teacher; the ViT-S (21M) variant is adopted as the per-individual embedder. On the Edinburgh Pig dataset, the compressed pipeline reaches 92.29% MOTA and 96.15% IDF1 against the SAM 3 teacher (1.68- and 0.84-percentage-point losses), achieves a 7.77-fold reduction in system-level parameters and a 3.01-fold reduction in peak VRAM (19.52GB -> 6.49GB), and reaches 97.34% top-1 accuracy with 91.67% macro-F1 on nine-class pig behaviour classification. The pipeline fits inside an NVIDIA Jetson Orin NX 16GB envelope with 4.9GB of headroom, supporting a proposed -- but not yet empirically validated -- on-device embedding-pool re-identification mechanism whose per-individual footprint of approximately 94MB per animal per year produces a longitudinal visual record amenable to retrospective association with disease, lameness, reproductive, and growth outcome labels.
Fonte: arXiv cs.CV
Vision • Score 85
EdgeFM: Efficient Edge Inference for Vision-Language Models
arXiv:2604.27476v1 Announce Type: new
Abstract: Vision-language models (VLMs) have demonstrated strong applicability in edge industrial applications, yet their deployment remains severely constrained by requirements for deterministic low latency and stable execution under resource limitations. Existing frameworks either rely on bloated general-purpose designs or force developers into opaque, hardware-specific closed-source ecosystems, leading to hardware lock-in limitation and poor cross-platform adaptability. Observing that modern AI agents can efficiently search and tune configurations to generate highly optimized low-level kernels for standard LLM operators, we propose EdgeFM, a lightweight, agent-driven VLM/LLM inference framework tailored for cross-platform industrial edge deployment. EdgeFM removes non-essential features to reduce single-request latency, and encapsulates agent-tuned kernel optimizations as a modular library of reusable skills. By allowing direct invocation of these skills rather than waiting for closed-source implementations, it effectively closes the performance gap long dominated by proprietary toolchains. The framework natively supports mainstream platforms including x86 and NVIDIA Orin SoCs, and represents the first end-to-end VLA deployment on the domestic Horizon Journey platform, enhancing cross-platform portability. In most cases, it yields clearly better inference performance than conventional vendor-specific toolchains, achieving up to 1.49 times speedup over TensorRT-Edge-LLM on the NVIDIA Orin platform. Experimental results show that EdgeFM delivers favorable end-to-end inference performance, providing an open-source, production-grade solution for diverse edge industrial scenarios.
Fonte: arXiv cs.CV
Vision • Score 85
World2Minecraft: Occupancy-Driven Simulated Scenes Construction
arXiv:2604.27578v1 Announce Type: new
Abstract: Embodied intelligence requires high-fidelity simulation environments to support perception and decision-making, yet existing platforms often suffer from data contamination and limited flexibility. To mitigate this, we propose World2Minecraft to convert real-world scenes into structured Minecraft environments based on 3D semantic occupancy prediction. In the reconstructed scenes, we can effortlessly perform downstream tasks such as Vision-Language Navigation(VLN). However, we observe that reconstruction quality heavily depends on accurate occupancy prediction, which remains limited by data scarcity and poor generalization in existing models. We introduce a low-cost, automated, and scalable data acquisition pipeline for creating customized occupancy datasets, and demonstrate its effectiveness through MinecraftOcc, a large-scale dataset featuring 100,165 images from 156 richly detailed indoor scenes. Extensive experiments show that our dataset provides a critical complement to existing datasets and poses a significant challenge to current SOTA methods. These findings contribute to improving occupancy prediction and highlight the value of World2Minecraft in providing a customizable and editable platform for personalized embodied AI research. Project page:https://world2minecraft.github.io/.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
People-Centred Medical Image Analysis
arXiv:2604.26991v1 Announce Type: new
Abstract: Recent advances in data-centric medical AI have produced highly accurate diagnostic systems, but the emphasis on data curation and performance metrics has not translated into widespread clinical adoption. We conjecture that this limited uptake stems from insufficient attention dedicated to the optimisation of fair performance across diverse patient populations and to workflow integration: performance biases can create regulatory barriers, and poorly integrated automation can disrupt clinical routines, degrade the quality of human-AI collaboration, and reduce clinicians' willingness to adopt AI tools. Prior work on workflow integration (e.g., Learning to Defer (L2D) and Learning to Complement (L2C)) and AI fairness has typically examined these challenges in isolation, overlooking their natural interdependence and the practical constraints of clinical environments, such as restricted clinician availability. We propose People-Centred Medical Image Analysis (PecMan), a human-AI framework that jointly optimises fairness, diagnostic accuracy, and workflow effectiveness through a dynamic gating mechanism that assigns cases to AI, clinicians, or both under clinician workload constraints. We also introduce the Fairness and Human-Centred AI (FairHAI) benchmark for evaluating trade-offs between accuracy, fairness, and clinician workload. Experiments using this benchmark show that PecMan consistently outperforms existing methods, paving the way for more trustworthy and clinically viable AI systems. Code will be available upon paper acceptance.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
YOSE: You Only Select Essential Tokens for Efficient DiT-based Video Object Removal
arXiv:2604.27322v1 Announce Type: new
Abstract: Recent advances in Diffusion Transformer (DiT)-based video generation technologies have shown impressive results for video object removal. However, these methods still suffer from substantial inference latency. For instance, although MiniMax Remover achieves state-of-the-art visual quality, it operates at only around 10FPS, primarily due to dense computations over the entire spatiotemporal token space, even when only a small masked region actually requires processing. In this paper, we present YOSE, You Only Select Essential Tokens, an efficient fine-tuning framework. YOSE introduces two key components: Batch Variable-length Indexing (BVI) and Diffusion Process Simulator (DiffSim) Module. BVI is a differentiable dynamic indexing operator that adaptively selects essential tokens based on mask information, enabling variable-length token processing across samples. DiffSim provides a diffusion process approximation mechanism for unmasked tokens, which simulates the influence of unmasked regions within DiT self-attention to maintain semantic consistency for masked tokens. With these designs, YOSE achieves mask-aware acceleration, where the inference time scales approximately linearly with the masked regions, in contrast to full-token diffusion methods whose computation remains constant regardless of the mask size. Extensive experiments demonstrate that YOSE achieves up to 2.5X speedup in 70% of cases while maintaining visual quality comparable to the baseline. Code is available at: https://github.com/Wucy0519/YOSE-CVPR26.
Fonte: arXiv cs.CV
Theory/Optimization • Score 85
Contextual Agentic Memory is a Memo, Not True Memory
arXiv:2604.27707v1 Announce Type: new
Abstract: Current agentic memory systems (vector stores, retrieval-augmented generation, scratchpads, and context-window management) do not implement memory: they implement lookup. We argue that treating lookup as memory is a category error with provable consequences for agent capability, long-term learning, and security. Retrieval generalizes by similarity to stored cases; weight-based memory generalizes by applying abstract rules to inputs never seen before. Conflating the two produces agents that accumulate notes indefinitely without developing expertise, face a provable generalization ceiling on compositionally novel tasks that no increase in context size or retrieval quality can overcome, and are structurally vulnerable to persistent memory poisoning as injected content propagates across all future sessions. Drawing on Complementary Learning Systems theory from neuroscience, we show that biological intelligence solved this problem by pairing fast hippocampal exemplar storage with slow neocortical weight consolidation, and that current AI agents implement only the first half. We formalize these limitations, address four alternative views, and close with a co-existence proposal and a call to action for system builders, benchmark designers, and the memory community.
Fonte: arXiv cs.AI
Privacy/Security/Fairness • Score 85
Automatic Causal Fairness Analysis with LLM-Generated Reporting
arXiv:2604.27011v1 Announce Type: new
Abstract: AutoML, intended as the process of automating the application of machine learning to real-world problems, is a key step for AI popularisation. Most AutoML frameworks are not accounting for the potential lack of fairness in the training data and in the corresponding predictions. We introduce \textsc{FairMind}, a software prototype aiming to automatise fairness analysis at the dataset level. We achieve that by resorting to the assumptions of the \emph{standard fairness model}, recently proposed by Ple\v{c}ko and Bareinboim. This allows for a sound fairness evaluation in terms of causal effects, based on \emph{counterfactual} queries involving the target, possibly confounders and mediators, and the different values of an input feature we regard as \emph{protected}. After the necessary data preprocessing, the tool implements a closed-form computation of the effects. LLMs are consequently exploited to generate accurate reports on the fairness levels detected in the training dataset. We achieve that in a zero-shot setup and show by examples the expected advantages with respect to a direct analysis performed by the LLM. To favour applications, extensions to ordinal protected variable and continuous targets and novel decomposition results are also discussed.
Fonte: arXiv cs.LG
MLOps/Systems • Score 88
TRUST: A Framework for Decentralized AI Service v.0.1
arXiv:2604.27132v1 Announce Type: new
Abstract: Large Reasoning Models (LRMs) and Multi-Agent Systems (MAS) in high-stakes domains demand reliable verification, yet centralized approaches suffer four limitations: (1) Robustness, with single points of failure vulnerable to attacks and bias; (2) Scalability, as reasoning complexity creates bottlenecks; (3) Opacity, as hidden auditing erodes trust; and (4) Privacy, as exposed reasoning traces risk model theft. We introduce TRUST (Transparent, Robust, and Unified Services for Trustworthy AI), a decentralized framework with three innovations: (i) Hierarchical Directed Acyclic Graphs (HDAGs) that decompose Chain-of-Thought reasoning into five abstraction levels for parallel distributed auditing; (ii) the DAAN protocol, which projects multi-agent interactions into Causal Interaction Graphs (CIGs) for deterministic root-cause attribution; and (iii) a multi-tier consensus mechanism among computational checkers, LLM evaluators, and human experts with stake-weighted voting that guarantees correctness under 30% adversarial participation. We prove a Safety-Profitability Theorem ensuring honest auditors profit while malicious actors incur losses. All decisions are recorded on-chain, while privacy-by-design segmentation prevents reconstruction of proprietary logic. Across multiple LLMs and benchmarks, TRUST attains 72.4% accuracy (4-18% above baselines) and remains resilient against 20% corruption. DAAN reaches 70% root-cause attribution (vs. 54-63% for standard methods) with 60% token savings. Human studies validate the design (F1 = 0.89, Brier = 0.074). The framework supports (A1) decentralized auditing, (A2) tamper-proof leaderboards, (A3) trustless data annotation, and (A4) governed autonomous agents, pioneering decentralized AI auditing for safe, accountable deployment of reasoning-capable systems.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Think it, Run it: Autonomous ML pipeline generation via self-healing multi-agent AI
arXiv:2604.27096v1 Announce Type: new
Abstract: The purpose of our paper is to develop a unified multi-agent architecture that automates end-to-end machine learning (ML) pipeline generation from datasets and natural-language (NL) goals, improving efficiency, robustness and explainability. A five-agent system is proposed to handle profiling, intent parsing, microservice recommendation, Directed Acyclic Graph (DAG) construction and execution. It integrates code-grounded Retrieval-Augmented Generation (RAG) for microservice understanding, an explainable hybrid recommender combining multiple criteria, a self-healing mechanism using Large Language Model (LLM)-based error interpretation and adaptive learning from execution history. The approach is evaluated on 150 ML tasks across diverse scenarios. The system achieves an 84.7% end-to-end pipeline success rate, outperforming baseline methods. It demonstrates improved robustness through self-healing and reduces workflow development time compared to manual construction. The study introduces a novel integration of code-grounded RAG, explainable recommendation, self-healing execution and adaptive learning within a single architecture, showing that tightly coupled intelligent components can outperform isolated solutions.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Belief-Guided Inference Control for Large Language Model Services via Verifiable Observations
arXiv:2604.27536v1 Announce Type: new
Abstract: In black-box large language model (LLM) services, response reliability is often only partially observable at decision time, while stronger inference pathways incur substantial computational cost, inducing a budgeted sequential decision problem: for each request, the system should decide whether the default low-cost response is sufficiently reliable or whether additional computation should be allocated to improve response quality. In this paper, we propose \textbf{Ver}ifiable \textbf{O}bservations for Risk-aware \textbf{I}nference \textbf{C}ontrol (\textsc{Veroic}), a framework for adaptive inference control in black-box LLM settings, which formulates request-time control as a \textit{partially observable Markov decision process} to capture partial observability and sequential budget coupling. It constructs a lightweight verifiable observation channel from the input-output pair by aggregating heterogeneous quality signals into a belief state over latent response reliability, which is then used by a budget-aware policy to decide whether to return the default output or trigger a higher-cost inference pathway. Experiments on diverse tasks show that \textsc{Veroic} achieves improved quality-cost trade-offs, stronger risk estimation and calibration, and more robust long-horizon inference control than competitive baselines.
Fonte: arXiv cs.AI
Vision • Score 85
Energy-Efficient Plant Monitoring via Knowledge Distillation
arXiv:2604.27178v1 Announce Type: new
Abstract: Recent advances in large-scale visual representation learning have significantly improved performance in plant species and plant disease recognition tasks. However, state-of-the-art models, often based on high-capacity vision transformers or multimodal foundation models, remain computationally expensive and difficult to deploy in resource-constrained environments such as mobile or edge devices. This limitation hinders the scalability of automated biodiversity monitoring and precision agriculture systems, where efficiency is as critical as accuracy. In this work, we investigate knowledge distillation as an effective approach to transfer the representational capacity of large pretrained models into smaller, more efficient architectures. We focus on plant species and disease recognition, and conduct an extensive empirical study on two challenging benchmarks: Pl@ntNet300K-v2 and Deep-Plant-Disease. We evaluate four representative architectures, including two ConvNeXt models and two vision transformers, under multiple training regimes: from-scratch training and pretrained initialization, each with and without distillation. In total, we train and evaluate 70 models. Our results show that knowledge distillation consistently improves performance across tasks and architectures. Distilled models are able to match the performance of significantly larger models while maintaining substantially lower computational cost. These findings demonstrate the potential of knowledge distillation techniques to enable efficient and scalable deployment of plant recognition systems in real-world environmental applications.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
Monitoring Neural Training with Topology: A Footprint-Predictable Collapse Index
arXiv:2604.26984v1 Announce Type: new
Abstract: Representational collapse, where embeddings become anisotropic and lose multi-scale structure, can erode downstream performance long before performance metrics react. We propose an online, topology-aware monitor for evolving neural representations that couples Modular Morse Homology Maintenance (MMHM) with a composite Collapse Index (CI). Instead of rebuilding complexes each epoch, we apply sparse edits at a fixed scale and maintain a discrete Morse matching, yielding fast, incremental updates. Across LLM fine-tuning and temporal KGE training, CI provides a low-latency early-warning signal suitable for in-training interventions. Code and experimental scripts will be released publicly
Fonte: arXiv cs.LG
MLOps/Systems • Score 90
End-to-End Evaluation and Governance of an EHR-Embedded AI Agent for Clinicians
arXiv:2604.27309v1 Announce Type: new
Abstract: Clinical AI systems require not just point-in-time evaluation but continuous governance: the ongoing practice of monitoring, evaluating, iterating, and re-evaluating performance throughout deployment. We present an end-to-end framework of governance that integrates rubric validation, live deployment feedback, technical performance monitoring, and cost tracking, with controlled experimentation gating system changes before deployment. Applied to Hyperscribe, an EHR-embedded agent that converts ambient audio into structured chart updates, twenty clinicians authored 1,646 validated rubrics across 823 cases. Seven Hyperscribe versions were evaluated through controlled experiments, with median scores improving from 84% to 95%. Analysis of 107 live feedback entries over three months showed feedback composition shifting from 79% error reports and 14% positive observations to 30% errors and 45% positive observations as engineering interventions resolved failures. Median processing time per audio segment was 8.1 seconds with a 99.6% effective completion rate after retry mechanisms absorbed transient model errors. These results demonstrate that continuous, multi-channel governance of deployed clinical AI is both achievable and effective.
Fonte: arXiv cs.AI
Privacy/Security/Fairness • Score 85
AdaBFL: Multi-Layer Defensive Adaptive Aggregation for Bzantine-Robust Federated Learning
arXiv:2604.27434v1 Announce Type: new
Abstract: Federated learning (FL) is a popular distributed learning paradigm in machine learning, which enables multiple clients to collaboratively train models under the guidance of a server without exposing private client data. However, FL's decentralized nature makes it vulnerable to poisoning attacks, where malicious clients can submit corrupted models to manipulate the system. To counter such attacks, although various Byzantine-robust methods have been proposed, these methods struggle to provide balanced defense against multiple types of attacks or rely on possessing the dataset in the server. To deal with these drawbacks, thus, we propose an effective multi-layer defensive adaptive aggregation for Bzantine-robust federated learning (AdaBFL) based on a novel three-layer defensive mechanism, which can adaptively adjust the weights of defense algorithms to counter complex attacks. Moreover, we provide convergence properties of our AdaBFL method under the non-convex setting on non-iid data. Comprehensive experiments across multiple datasets validate the superiority of our AdaBFL over the comparable algorithms.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Skills-Coach: A Self-Evolving Skill Optimizer via Training-Free GRPO
arXiv:2604.27488v1 Announce Type: new
Abstract: We introduce Skills-Coach, a novel automated framework designed to significantly enhance the self-evolution of skills within Large Language Model (LLM)-based agents. Addressing the current fragmentation of the skill ecosystem, Skills-Coach explores the boundaries of skill capabilities, thereby facilitating the comprehensive competency coverage essential for intelligent applications. The framework comprises four core modules: a Diverse Task Generation Module that systematically creates a comprehensive test suite for various skills; a Lightweight Optimization Module dedicated to optimizing skill prompts and their corresponding code; a Comparative Execution Module facilitating the execution and evaluation of both original and optimized skills; and a Traceable Evaluation Module, which rigorously evaluates performance against specified criteria. Skills-Coach offers flexible execution options through its virtual and real modes. To validate its efficacy, we introduce Skill-X, a comprehensive benchmark dataset consisting of 48 diverse skills. Experimental results demonstrate that Skills-Coach achieves significant performance improvements in skill capability across a wide range of categories, highlighting its potential to advance the development of more robust and adaptable LLM-based agents.
Fonte: arXiv cs.CL
MLOps/Systems • Score 85
Toward Scalable SDN for LEO Mega-Constellations: A Graph Learning Approach
arXiv:2604.27478v1 Announce Type: new
Abstract: Terrestrial network limitations drive the integration of non-terrestrial networks (NTNs), notably mega-constellations comprising thousands of low Earth orbit (LEO) satellites. While these satellites act as interconnected network switches via inter-satellite links (ISLs), their massive scale creates severe bottlenecks for network management. To address this, we propose a scalable, hierarchical software-defined networking (SDN) framework. Our architecture leverages graph neural networks (GNNs) to compactly represent the constellation topology, and Koopman theory to linearize nonlinear dynamics. Specifically, a Graph Koopman Autoencoder (GKAE) forecasts spatio-temporal behavior within a linear subspace for each orbital shell. A central SDN controller then aggregates these shell-level predictions for globally coordinated control. Simulations on the Starlink constellation demonstrate that our approach achieves at least a 42.8\% improvement in spatial compression and a 10.81\% improvement in temporal forecasting compared to established baselines, all while utilizing a significantly smaller model footprint.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
The Polynomial Stein Discrepancy for Assessing Moment Convergence
arXiv:2412.05135v2 Announce Type: replace
Abstract: We propose a novel method for measuring the discrepancy between a set of samples and a desired posterior distribution for Bayesian inference. Classical methods for assessing sample quality like the effective sample size are not appropriate for scalable Bayesian sampling algorithms, such as stochastic gradient Langevin dynamics, that are asymptotically biased. Instead, the gold standard is to use the kernel Stein Discrepancy (KSD), which is itself not scalable given its quadratic cost in the number of samples. The KSD and its faster extensions also typically suffer from the curse of dimensionality and can require extensive tuning. To address these limitations, we develop the polynomial Stein discrepancy (PSD) and an associated goodness-of-fit test. While the new test is not fully convergence-determining, we prove that it detects differences in the first r moments for Gaussian targets. We empirically show that the test has higher power than its competitors in several examples, and at a lower computational cost. Finally, we demonstrate that the PSD can assist practitioners to select hyper-parameters of Bayesian sampling algorithms more efficiently than competitors.
Fonte: arXiv stat.ML
Vision • Score 85
Automated Detection of Mutual Gaze and Joint Attention in Dual-Camera Settings via Dual-Stream Transformers
arXiv:2604.27105v1 Announce Type: new
Abstract: Analyzing mutual gaze (MG) and joint attention (JA) is critical in developmental psychology but traditionally relies on labor-intensive manual coding. Automating this process in multi-camera laboratory settings is computationally challenging due to complex cross-camera relational dynamics. In this paper, we propose a highly efficient dual-stream Transformer architecture for detecting MG and JA from synchronized dual-camera recordings. Our approach leverages frozen gaze-aware backbones (GazeLLE) to extract rich visual priors, combined with a custom token fusion mechanism to map the spatial and semantic relationships between interacting dyads. Evaluated on an ecologically valid dataset of caregiver-infant interactions, our model exhibits good performance, significantly outperforming both a convolutional baseline and a state-of-the-art multimodal Large Language Model (LLM). By open-sourcing our model and pre-trained weights, we provide behavioral scientists with a scalable tool that can be fine-tuned to diverse laboratory environments, effectively bridging the gap between computational modeling and applied interaction research.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
SCOPE-FE: Structured Control of Operator and Pairwise Exploration for Feature Engineering
arXiv:2604.27025v1 Announce Type: new
Abstract: Automatic feature engineering is an effective approach for improving predictive performance in tabular learning. However, expand-and-reduce methods, such as OpenFE, become increasingly computationally expensive as the input dimensionality grows. This limitation arises primarily from the combinatorial explosion of candidate features generated through operator-feature combinations. To address this issue, we propose SCOPE-FE, a structured search space control framework that improves efficiency by reducing the candidate space prior to feature generation. SCOPE-FE jointly regulates two major sources of combinatorial growth: the operator space and feature-pair space. First, OperatorProbing estimates the dataset-specific utility of candidate operators and eliminates low-contribution operators in advance. Second, FeatureClustering employs spectral embedding and fuzzy c-means clustering to group structurally related features, thereby restricting candidate generation to relevant within-cluster combinations. In addition, we introduce ReliabilityScoring, which incorporates variance across subsamples to stabilize pruning decisions. Experiments on ten benchmark datasets demonstrate that SCOPE-FE substantially reduces feature engineering time while maintaining competitive predictive performance relative to existing baselines. The efficiency gains are particularly pronounced for high-dimensional datasets. These results indicate that structured control of the search space is an effective strategy for scalable automatic feature engineering. The code will be made publicly available upon acceptance.
Fonte: arXiv stat.ML
RL • Score 85
Optimal Stop-Loss and Take-Profit Parameterization for Autonomous Trading Agent Swarm
arXiv:2604.27150v1 Announce Type: new
Abstract: Autonomous crypto trading systems often spend most of their design effort on finding entries, while exits are left to fixed rules that are rarely tested in a systematic way. This paper examines whether better stop-loss and take-profit settings can improve the performance of an autonomous trading agent swarm. Using more than 900 historical trades, we replay each trade under many alternative exit policies and compare results against the existing production setup. The study finds that exit design matters meaningfully: stronger configurations improve risk-adjusted performance and generally favor tighter loss limits, earlier profit capture, and closer trailing protection. The paper also discusses a key evaluation challenge: a purely chronological split was initially used, but the newest trades fell into an unusual war-driven market period that sharply distorted test results. To reduce the influence of that single episode, the main comparison was run on randomized data, with the drawbacks of doing so acknowledged explicitly. Overall, the paper presents a practical framework for tuning exit logic in a more disciplined and transparent way.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
NORACL: Neurogenesis for Oracle-free Resource-Adaptive Continual Learning
arXiv:2604.27031v1 Announce Type: new
Abstract: In a continual learning setting, we require a model to be plastic enough to learn a new task and stable enough to not disturb previously learned capabilities. We argue that this dilemma has an architectural root. A finite network has limited representational and plastic resources, yet the required capacity depends on properties of the future task stream that are unknown: how many tasks will be encountered, and how much they overlap in feature space. Regularization-based methods preserve past knowledge within fixed-capacity architectures and therefore implicitly rely on an oracle architecture sized for this unknown future. When tasks are only weakly related, fixed architectures progressively run out of plastic resources; when tasks are few or strongly overlapping, models are often over-provisioned. Inspired by neurogenesis in biology, we propose NORACL to address the stability-plasticity dilemma by tackling the oracle architecture problem through neuronal growth. Starting from a compact network, NORACL grows only when needed by monitoring two complementary signals for representational and plasticity saturation. We evaluate NORACL against oracle-sized static baselines across varying task counts and geometries. Across all settings, NORACL achieves final average accuracies that are better than or on par with oracle-provisioned static baselines while using fewer parameters. Additionally, NORACL yields architectures with interpretable growth, i.e. dissimilar tasks predominantly expand feature-extraction layers, whereas tasks which rely on common features shift growth toward later feature-combination layers. Our analysis further explains why fixed-capacity networks lose plasticity as tasks accumulate, whereas NORACL creates fresh capacity for new tasks through growth. Together, these results show that adaptive neurogenesis pushes the stability-plasticity Pareto frontier of continual learning.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
FMCL: Class-Aware Client Clustering with Foundation Model Representations for Heterogeneous Federated Learning
arXiv:2604.27510v1 Announce Type: new
Abstract: Federated Learning (FL) enables collaborative model training across distributed clients without sharing raw data, yet its performance deteriorates under statistical heterogeneity. Clustered Federated Learning addresses this challenge by grouping similar clients and training separate models per cluster. However, existing clustering strategies often rely on raw data statistics, model parameters, or heuristic similarity measures that fail to capture class-level semantic structure across heterogeneous domains and frequently require iterative coordination. We propose FMCL, a one-shot, class-aware client clustering framework that leverages foundation model representations to construct semantic client signatures. Using a frozen foundation model, FMCL computes class-level embedding prototypes for each client and measures similarity via cosine distance between their class-aware representations. Clustering is performed once prior to training, introducing no additional communication during federated optimization and remaining agnostic to the downstream model architecture. Extensive experiments across heterogeneous benchmarks demonstrate that FMCL improves federated performance and yields more stable clustering behavior compared to existing clustering-based methods under non-identically distributed data partitioning.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
General Uncertainty Estimation with Delta Variances
arXiv:2502.14698v2 Announce Type: replace-cross
Abstract: Decision makers may suffer from uncertainty induced by limited data. This may be mitigated by accounting for epistemic uncertainty, which is however challenging to estimate efficiently for large neural networks. To this extent we investigate Delta Variances, a family of algorithms for epistemic uncertainty quantification, that is computationally efficient and convenient to implement. It can be applied to neural networks and more general functions composed of neural networks. As an example we consider a weather simulator with a neural-network-based step function inside -- here Delta Variances empirically obtain competitive results at the cost of a single gradient computation. The approach is convenient as it requires no changes to the neural network architecture or training procedure. We discuss multiple ways to derive Delta Variances theoretically noting that special cases recover popular techniques and present a unified perspective on multiple related methods. Finally we observe that this general perspective gives rise to a natural extension and empirically show its benefit.
Fonte: arXiv stat.ML
Vision • Score 85
Exploring Vision Neural Network Pruning via Screening Methodology
arXiv:2502.07189v2 Announce Type: replace-cross
Abstract: The remarkable performance of modern deep neural networks (DNNs) is largely driven by their massive scale, often comprising tens to hundreds of millions-or even billions-of parameters. However, such a scale incurs substantial storage and computational costs, hindering deployment on platforms such as edge devices that require energy-efficient and real-time processing. In this paper, we propose a network pruning framework that reduces both storage and computation requirements by an order of magnitude while preserving model accuracy. Our approach eliminates non-essential parameters through a statistical analysis of component significance across classification categories. Specifically, we employ a F-statistic-based screening technique combined with a weighted evaluation scheme to quantify the contributions of connections and channels, enabling both unstructured and structured pruning within a unified framework. Extensive experiments on real-world vision datasets, covering both fully connected neural networks (FNNs) and convolutional neural networks (CNNs), demonstrate that the proposed framework produces compact and efficient models that are highly competitive with the state of art apporoaches.
Fonte: arXiv stat.ML
MLOps/Systems • Score 85
DeepWeightFlow: Re-Basined Flow Matching for Generating Neural Network Weights
arXiv:2601.05052v2 Announce Type: replace-cross
Abstract: Building efficient and effective generative models for neural network weights has been a research focus of significant interest that faces challenges posed by the high-dimensional weight spaces of modern neural networks and their symmetries. Several prior generative models are limited to generating partial neural network weights, particularly for larger models, such as ResNet and ViT. Those that do generate complete weights struggle with generation speed or require finetuning of the generated models. In this work, we present DeepWeightFlow, a Flow Matching model that operates directly in weight space to generate diverse and high-accuracy neural network weights for a variety of architectures, neural network sizes, and data modalities. The neural networks generated by DeepWeightFlow do not require fine-tuning to perform well and can scale to large networks. We apply Git Re-Basin and TransFusion for neural network canonicalization in the context of generative weight models to account for the impact of neural network permutation symmetries and to improve generation efficiency for larger model sizes. The generated networks excel at transfer learning, and ensembles of hundreds of neural networks can be generated in minutes, far exceeding the efficiency of diffusion-based methods. DeepWeightFlow models pave the way for more efficient and scalable generation of diverse sets of neural networks.
Fonte: arXiv stat.ML
MLOps/Systems • Score 85
FoReco and FoRecoML: A Unified Toolbox for Forecast Reconciliation in R
arXiv:2604.27696v1 Announce Type: cross
Abstract: Forecast reconciliation has become key to improving the accuracy and coherence of forecasts for linearly constrained multiple time series, such as hierarchical and grouped series. Yet, comprehensive software that jointly covers cross-sectional, temporal, and cross-temporal reconciliation has so far been lacking. The R packages FoReco and FoRecoML address this gap by offering a comprehensive and unified framework. The packages respectively implement classical and regression-based linear reconciliation approaches, and non-linear approaches based on machine learning for cross-sectional, temporal and cross-temporal frameworks. Designed for accessibility and flexibility, these packages provide sensible default options that allow new users to apply reconciliation methods with minimal effort, while still giving expert users full control to explore state-of-the-art extensions through customized settings. With this dual focus, FoReco and FoRecoML are versatile tools for practitioners and researchers working on forecast reconciliation.
Fonte: arXiv stat.ML
Privacy/Security/Fairness • Score 85
Privacy-Preserving Federated Learning via Differential Privacy and Homomorphic Encryption for Cardiovascular Disease Risk Modeling
arXiv:2604.27598v1 Announce Type: new
Abstract: Protecting sensitive health data while enabling collaborative analysis is a central challenge in healthcare. Traditional machine learning (ML) requires institutions to pool anonymized patient records, centralizing analytical development and privacy risks at a single site. Privacy-enhancing technologies (PETs), including Differential Privacy (DP) and Homomorphic Encryption (HE), can mitigate these risks. However, they are mainly studied in conventional data-sharing settings and often introduce trade-offs, including reduced model utility, higher computational cost, and increased implementation complexity. Federated Learning (FL) reduces data centralization by enabling institutions to train models locally and share only model updates. Nevertheless, FL does not eliminate privacy risks, as shared parameters or gradients may still reveal sensitive information. Integrating DP or HE into FL can strengthen privacy guarantees, yet their comparative performance and deployment implications in real-world healthcare settings remain unclear.
We systematically evaluated DP and HE integration in FL under real-world conditions, comparing them with standard FL and centralized ML (cML) to quantify privacy-utility trade-offs in multi-institutional settings. Using nationwide Swedish healthcare data, we evaluated cardiovascular disease risk prediction using logistic regression (LR) and neural network (NN) learners. FL with HE achieved performance comparable to cML but introduced measurable cryptographic overhead, particularly in the NN implementation. FL with DP incurred lower computational cost; however, LR was more sensitive to calibrated noise than the NN, resulting in greater performance degradation. Our findings provide practical guidance for deploying privacy-preserving FL in fragmented healthcare systems.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
The Bernstein-von Mises theorem for Bayesian one-pass online learning
arXiv:2604.27442v1 Announce Type: cross
Abstract: Bayesian online learning provides a coherent framework for sequential inference. However, its theoretical understanding remains limited, particularly in the one-pass setting. Existing theoretical guarantees typically require the mini-batch sample size to diverge, a condition that fails in the one-pass regime. In this paper, we propose a new Bayesian online learning algorithm tailored to the one-pass setting, which incorporates a warm-start phase to ensure stable sequential updates. For this algorithm, we show that the sequentially updated posterior attains the optimal convergence rate. Building on this, we establish an online analogue of the Bernstein-von Mises theorem, which guarantees valid uncertainty quantification without diverging mini-batch sample sizes. Our analysis is based on a novel theoretical framework that differs fundamentally from existing approaches in the online learning literature. Numerical experiments on generalized linear models show that the proposed method matches the performance of the batch estimator while outperforming existing online procedures.
Fonte: arXiv stat.ML
Theory/Optimization • Score 85
Statistical Channel Fingerprint Construction for Massive MIMO: A Unified Tensor Learning Framework
arXiv:2604.27574v1 Announce Type: new
Abstract: Channel fingerprint (CF) is considered a key enabler for facilitating the acquisition of channel state information (CSI) in massive multiple-input multiple-output (MIMO) communication systems. In this work, we investigate a novel type of CF that stores statistical CSI (sCSI) at each potential location, referred to as statistical CF (sCF). Specifically, we reveal the relationship between sCSI, namely the channel spatial covariance matrix (CSCM), and the channel power angular spectrum (CPAS). Building on this foundation, we construct a unified tensor representation of the sCF and further reduce its dimension by exploiting the eigenvalue decomposition of the CSCM and its correlation with the PAS. Considering the practical constraints imposed by measurement cost, privacy, and security, we focus on three representative scenarios and uniformly formulate them as tensor restoration tasks. To this end, we propose a unified tensor-based learning architecture, termed LPWTNet. The architecture incorporates a closed-form Laplacian pyramid (LP) decomposition and reconstruction framework that replaces the traditional encoder-decoder structure, enabling efficient inference while capturing multi-scale frequency subband characteristics of the sCF. Additionally, a shared mask learning strategy is introduced to adaptively refine high-frequency sCF components through level-wise adjustments. To achieve a larger receptive field without over-parameterization, we further propose a small-kernel convolution mechanism based on the wavelet transform (WT), which decouples convolution across different frequency components of the sCF and enhances feature extraction efficiency. Extensive experiments show that the proposed approach delivers competitive reconstruction accuracy and computational efficiency across various sCF construction scenarios when compared with state-of-the-art baselines.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
TypeBandit: Type-Level Context Allocation and Reweighting for Effective Attribute Completion in Heterogeneous Graph Neural Networks
arXiv:2604.27356v1 Announce Type: new
Abstract: Heterogeneous graphs are widely used to model multi-relational systems, but missing node attributes remain a major bottleneck for downstream learning. In this paper, we identify and formalize type-dependent information asymmetry: the phenomenon that different node types provide substantially different levels of useful signal for attribute completion. Motivated by this observation, we propose TypeBandit, a lightweight, model-agnostic methodology for heterogeneous attribute completion. TypeBandit combines topology-aware initialization, type-level bandit sampling, and joint representation learning. It allocates a finite global sampling budget across node types, samples representative nodes within each type, and uses the resulting sampled type summaries as shared contextual signals during representation construction. By operating at the type level rather than over each target node's local neighborhood, TypeBandit keeps the adaptive state compact and practical for large heterogeneous graphs.
A key advantage of TypeBandit is architectural flexibility. Rather than requiring a new heterogeneous graph neural network architecture, TypeBandit acts as a type-aware front end for representative heterogeneous GNN backbones, including R-GCN, HetGNN, HGT, and SimpleHGN. We further introduce a hybrid pretraining scheme that combines structural degree priors with feature propagation, yielding a more reliable initializer than degree-only pretraining. Under a fixed-split protocol on DBLP, IMDB, and ACM, TypeBandit provides dataset-dependent but practically meaningful gains. Additional ablation, stability, efficiency, semantic-propagation, and sampled OGBN-MAG experiments support TypeBandit as a practical strategy for heterogeneous attribute completion when type-specific information is unevenly distributed and sampling resources are limited.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
ChipLingo: A Systematic Training Framework for Large Language Models in EDA
arXiv:2604.27415v1 Announce Type: new
Abstract: With the rapid advancement of semiconductor technology, Electronic Design Automation (EDA) has become an increasingly knowledge-intensive and document-driven engineering domain. Although large language models (LLMs) have shown strong general capabilities, applying them directly to EDA remains challenging due to limited domain expertise, cross-tool knowledge confusion, and degraded retrieval-augmented generation (RAG) performance after domain training. To address these issues, this paper presents ChipLingo, a systematic training pipeline for domain-adapted LLMs tailored to EDA scenarios.
ChipLingo consists of three stages: domain corpus construction with multi-source data curation and QA augmentation, domain-adaptive pretraining with comparisons of different parameter training strategies, and instruction alignment with RAG scenario training under diverse retrieval conditions. We also curate an internal benchmark, EDA-Bench, covering representative EDA tool scenarios, with plans for public release.
Experiments show that ChipLingo-8B achieves 59.7% accuracy on EDA-Bench, outperforming the same-scale base model and some larger general-purpose models. ChipLingo-32B reaches 70.02%, approaching leading closed-source commercial models. Further analysis shows that QA augmentation improves domain performance, Partial FT offers a better balance between adaptation and general capability retention than LoRA, and explicit RAG scenario training mitigates the decline in retrieval utilization after domain training. These results demonstrate the practical value of systematic domain training for knowledge-intensive EDA tasks and provide a foundation for future EDA agents and external-knowledge-driven systems.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Online semi-supervised perception: Real-time learning without explicit feedback
arXiv:2604.27562v1 Announce Type: new
Abstract: This paper proposes an algorithm for real-time learning without explicit feedback. The algorithm combines the ideas of semi-supervised learning on graphs and online learning. In particular, it iteratively builds a graphical representation of its world and updates it with observed examples. Labeled examples constitute the initial bias of the algorithm and are provided offline, and a stream of unlabeled examples is collected online to update this bias. We motivate the algorithm, discuss how to implement it efficiently, prove a regret bound on the quality of its solutions, and apply it to the problem of real-time face recognition. Our recognizer runs in real time, and achieves superior precision and recall on 3 challenging video datasets.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
A Short Note on Batch-efficient Divide-and-Conquer Algorithm for EigenDecomposition
arXiv:2604.27325v1 Announce Type: new
Abstract: EigenDecomposition (ED) is at the heart of many computer vision algorithms and applications. One crucial bottleneck limiting its usage is the expensive computation cost, particularly for a mini-batch of matrices in deep neural networks. Our previous work proposed a dedicated QR-based ED algorithm for batched small matrices (dim${<}32$). This short paper targets the limitation and proposes a batch-efficient Divide-and-Conquer based ED algorithm for larger matrices. The numerical test shows that for a mini-batch of matrices whose dimensions are smaller than $64$, our method can be much faster than the Pytorch SVD function.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
When Continual Learning Moves to Memory: A Study of Experience Reuse in LLM Agents
arXiv:2604.27003v1 Announce Type: new
Abstract: Memory-augmented LLM agents offer an appealing shortcut to continual learning: rather than updating model parameters, they accumulate experience in external memory, seemingly sidestepping the stability-plasticity dilemma of parametric learning. We show that this challenge does not disappear but resurfaces at the memory level. Under a limited context window, old and new experiences compete during retrieval, relocating the continual-learning bottleneck from parameter updates to memory access. To study this phenomenon, we introduce a (k,v) framework that disentangles two fundamental design axes of external memory: how experience is represented and how it is organized for retrieval. Across sequential-task experiments in ALFWorld and BabyAI, we find that abstract procedural memories transfer more reliably than detailed trajectories, while negative transfer disproportionately harms the hard cases. Moreover, finer-grained memory organization is not universally beneficial: designs that yield strong forward transfer can simultaneously induce severe forgetting. Together, these results reveal that external memory does not resolve the continual-learning problem; it reshapes it into a problem of memory representation and retrieval design.
Fonte: arXiv cs.LG
RL • Score 85
AutoREC: A software platform for developing reinforcement learning agents for equivalent circuit model generation from electrochemical impedance spectroscopy data
arXiv:2604.27266v1 Announce Type: new
Abstract: This paper introduces AutoREC, an open-source Python package for developing reinforcement learning (RL) agents to automatically generate equivalent circuit models (ECMs) from electrochemical impedance spectroscopy (EIS) data. While ECMs are a standard framework for interpreting EIS data, traditional identification is typically based on manual trial-and-error, which requires domain experts and limits scalability, particularly in autonomous experimental pipelines such as self-driving laboratories. AutoREC addresses this challenge by formulating ECM construction as a sequential decision-making problem within a Markov Decision Process framework. It implements a Double Deep Q-Network with prioritized experience replay, along with a dedicated dead-loop mitigation strategy, to efficiently explore a complex action space for circuit generation. To demonstrate the capabilities of the platform, we trained an RL agent using AutoREC and evaluated its strengths and limitations across diverse datasets, while also discussing possible strategies to mitigate these limitations in future agent designs. The trained agent achieved a success rate exceeding $99.6\%$ on synthetic datasets and demonstrated strong generalization to unseen experimental EIS data from batteries, corrosion, oxygen evolution reaction, and CO$_2$ reduction systems. These results position AutoREC as a promising platform for adaptive and data-driven ECM generation, with potential for integration into automated electrochemical workflows.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Generalizing the Geometry of Model Merging Through Frechet Averages
arXiv:2604.27155v1 Announce Type: new
Abstract: Model merging aims to combine multiple models into one without additional training. Na\"ive parameter-space averaging can be fragile under architectural symmetries, as their geometry does not take them into account. In this work we show that not only the geometry, but also the averaging procedure itself, must be symmetry-invariant to achieve symmetry-aware merges. Consequently, we propose a general solution: merging as Fr\'echet averaging, i.e., selecting parameters that minimize a sum of geodesic distances on an appropriate manifold. In this view, the key design choice is the overall geometry, i.e., the choice of metric, manifold, and distance approximation, that determines what it means for two models to be "close". We show that Fr\'echet averaging, combined with simplifying assumptions, contains Fisher merging. Building on this, we examine the particular case of low-rank adapters (LoRA), whose symmetries induce a distinct geometry: that of a quotient manifold. We outline the limitations of current LoRA merging methods, propose a practical algorithm for this setting, and show how they compare with other commonly used approaches.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
When Your LLM Reaches End-of-Life: A Framework for Confident Model Migration in Production Systems
arXiv:2604.27082v1 Announce Type: new
Abstract: We present a framework for migrating production Large Language Model (LLM) based systems when the underlying model reaches end-of-life or requires replacement. The key contribution is a Bayesian statistical approach that calibrates automated evaluation metrics against human judgments, enabling confident model comparison even with limited manual evaluation data. We demonstrate this framework on a commercial question-answering system serving 5.3M monthly interactions across six global regions; evaluating correctness, refusal behavior, and stylistic adherence to successfully identify suitable replacement models. The framework is broadly applicable to any enterprise deploying LLM-based products, providing a principled, reproducible methodology for model migration that balances quality assurance with evaluation efficiency. This is a capability increasingly essential as the LLM ecosystem continues to evolve rapidly and organizations manage portfolios of AI-powered services across multiple models, regions, and use cases.
Fonte: arXiv cs.AI
NLP/LLMs • Score 90
Better Models, Faster Training: Sigmoid Attention for single-cell Foundation Models
arXiv:2604.27124v1 Announce Type: new
Abstract: Training stable biological foundation models requires rethinking attention mechanisms: we find that using sigmoid attention as a drop in replacement for softmax attention a) produces better learned representations: on six diverse single-cell datasets, sigmoid achieves 25% higher cell-type separation, better cell-type cohesion metrics, and lower validation loss, b) faster training, models with sigmoid attention train up to 10% faster than their softmax counterparts, and c) more stable training by eliminating inherent sources of instability in softmax attention. We establish that sigmoid attention has globally bounded derivatives ($\leq 0.25$) as opposed to softmax, and a diagonal Jacobian structure in contrast with softmax's dense coupling, which together help alleviate training instabilities. In stress tests on 160M-parameter bidirectional attention models trained without gradient clipping on 8K-token sequences, softmax diverges catastrophically, with gradients exploding by four orders of magnitude, while sigmoid remains stable. Finally, we implement and open-source TritonSigmoid, an efficient GPU kernel that achieves 515 TFLOPS on H100 GPUs, outperforming both FlashAttention-2 and FlashSigmoid, with native padding support, which is essential for biological sequences. Our results establish sigmoid attention as both theoretically grounded and empirically superior for biological foundation models. Code is available at https://github.com/MSDLLCpapers/triton-sigmoid
Fonte: arXiv cs.LG
RL • Score 85
Co-Evolving Policy Distillation
arXiv:2604.27083v1 Announce Type: new
Abstract: RLVR and OPD have become standard paradigms for post-training. We provide a unified analysis of these two paradigms in consolidating multiple expert capabilities into a single model, identifying capability loss in different ways: mixed RLVR suffers from inter-capability divergence cost, while the pipeline of first training experts and then performing OPD, though avoiding divergence, fails to fully absorb teacher capabilities due to large behavioral pattern gaps between teacher and student. We propose Co-Evolving Policy Distillation (CoPD), which encourages parallel training of experts and introduces OPD during each expert's ongoing RLVR training rather than after complete expert training, with experts serving as mutual teachers (making OPD bidirectional) to co-evolve. This enables more consistent behavioral patterns among experts while maintaining sufficient complementary knowledge throughout. Experiments validate that CoPD achieves all-in-one integration of text, image, and video reasoning capabilities, significantly outperforming strong baselines such as mixed RLVR and MOPD, and even surpassing domain-specific experts. The model parallel training pattern offered by CoPD may inspire a novel training scaling paradigm.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Unsupervised Electrofacies Classification and Porosity Characterization in the Offshore Keta Basin Using Wireline Logs
arXiv:2604.27126v1 Announce Type: new
Abstract: This study presents an unsupervised machine learning workflow for electrofacies analysis in the offshore Keta Basin, Ghana, where core data are scarce. Six standard wireline logs from Well~C were analysed over a depth interval comprising approximately $11{,}195$ samples. K-means clustering was applied in multivariate log space, with the clustering structure evaluated using inertia and silhouette diagnostics. Four clusters were identified, supported by an average silhouette coefficient of approximately $0.50$, indicating moderate but meaningful separation. The resulting electrofacies exhibit systematic, depth-continuous patterns associated with variations in clay content, porosity, and rock framework properties, forming a geological continuum from shale-dominated to cleaner sandstone-dominated units. The results demonstrate that log-only, unsupervised clustering supported by quantitative metrics provides a robust and reproducible framework for subsurface characterisation. The proposed workflow offers a practical tool for early-stage formation evaluation in frontier offshore basins and a foundation for future integrated studies.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
WaferSAGE: Large Language Model-Powered Wafer Defect Analysis via Synthetic Data Generation and Rubric-Guided Reinforcement Learning
arXiv:2604.27629v1 Announce Type: new
Abstract: We present WaferSAGE, a framework for wafer defect visual question answering using small vision-language models. To address data scarcity in semiconductor manufacturing, we propose a three-stage synthesis pipeline incorporating structured rubric generation for precise evaluation. Starting from limited labeled wafer maps, we employ clustering-based cleaning to filter label noise, then generate comprehensive defect descriptions using vision-language models, which are converted into structured evaluation rubrics criteria. These rubrics guide the synthesis of VQA pairs, ensuring coverage across defect type identification, spatial distribution, morphology, and root cause analysis.
Our dual assessment framework aligns rule-based metrics with LLM-Judge scores via Bayesian optimization, enabling reliable automated evaluation. Through curriculum-based reinforcement learning with Group Sequence Policy Optimization (GSPO) and rubric-aligned rewards, our 4B-parameter Qwen3-VL model achieves a 6.493 LLM-Judge score, closely approaching Gemini-3-Flash (7.149) while enabling complete on-premise deployment. We demonstrate that small models with domain-specific training can surpass proprietary large models in specialized industrial visual understanding, offering a viable path for privacy-preserving, cost-effective deployment in semiconductor manufacturing.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Step-level Optimization for Efficient Computer-use Agents
arXiv:2604.27151v1 Announce Type: new
Abstract: Computer-use agents provide a promising path toward general software automation because they can interact directly with arbitrary graphical user interfaces instead of relying on brittle, application-specific integrations. Despite recent advances in benchmark performance, strong computer-use agents remain expensive and slow in practice, since most systems invoke large multimodal models at nearly every interaction step. We argue that this uniform allocation of compute is fundamentally inefficient for long-horizon GUI tasks. Such trajectories are highly heterogeneous: many steps are routine and can be handled reliably by smaller, cheaper policies, while errors tend to concentrate at a relatively small number of high-risk moments. Across computer-use benchmarks, these failures repeatedly take two forms: progress stalls, where the agent loops, repeats ineffective actions, or fails to make meaningful progress, and silent semantic drift, where the agent continues taking locally plausible actions after already deviating from the user's true goal. To address this inefficiency, we propose an event-driven, step-level cascade for computer-use agents that runs a small policy by default and escalates to a stronger model only when lightweight learned monitors detect elevated risk. Our framework combines two complementary signals: a Stuck Monitor that detects degraded progress from recent reasoning-action history and triggers recovery, and a Milestone Monitor that identifies semantically meaningful checkpoints where sparse verification is most informative for catching drift. This design turns always-on frontier-model inference into adaptive, on-demand compute allocation over the course of an evolving interaction. The framework is modular and deployment-oriented: it can be layered on top of existing computer-use agents without changing the underlying agent architecture or retraining the large model.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
The Inverse-Wisdom Law: Architectural Tribalism and the Consensus Paradox in Agentic Swarms
arXiv:2604.27274v1 Announce Type: new
Abstract: As AI transitions toward multi-agent systems (MAS) to solve complex workflows, research paradigms operate on the axiomatic assumption that agent collaboration mirrors the "Wisdom of the Crowd". We challenge this assumption by formalizing the Consensus Paradox: a phenomenon where agentic swarms prioritize internal architectural agreement over external logical truth. Through a 36 experiments encompassing 12,804 trajectories across three state-of-the-art (SOTA) benchmarks (GAIA, Multi-Challenge, and SWE-bench), we prove the Inverse-Wisdom Law: in kinship-dominant swarms, adding logical agents increases the stability of erroneous trajectories rather than the probability of truth. The introduction of additional logical audits converges the system toward a Logic Saturation where internal entropy hits zero while factual error hits unity. By evaluating the interaction between the 3 preeminent SOTA models (Gemini 3.1 Pro, Claude Sonnet 4.6, and GPT-5.4), we establish the Architectural Tribalism Asymmetry as a mechanistic law of transformer weights. We demonstrate that terminal swarm integrity is strictly gated by the synthesizer's receptive logic, rather than aggregate agent quality. We define the Tribalism Coefficient and the Sycophantic Weight as the primary mechanistic determinants of swarm failure. Finally, we establish the Heterogeneity Mandate as a foundational safety requirement for resilient agentic architectures.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
AutoSP: Unlocking Long-Context LLM Training Via Compiler-Based Sequence Parallelism
arXiv:2604.27089v1 Announce Type: new
Abstract: Large-language-models (LLMs) demonstrate enormous utility in long-context tasks which require processing prompts that consist of tens to hundreds of thousands of tokens. However, existing LLM training libraries do not provide easy to use abstractions to optimize for long-context training, instead focusing on optimizations for models with large parameter counts through ZeRO-3/FSDP, Tensor and Pipeline parallelism. This forces users to rewrite LLM training libraries to incorporate compositions of various complex long-context optimizations, such as sequence-parallelism, to training pipelines; a process that requires in-depth expertise, reducing developer productivity. To tackle these challenges, we introduce AutoSP: the first automated solution to automatically optimize LLM training for longer-contexts. AutoSP compiles models and applies a targeted set of optimizations: automated sequence parallelism, and long-context aware activation-checkpointing, to drastically enhance LLM trainability at negligible cost to throughput. Our evaluation demonstrates AutoSP's capability on both NVIDIA and AMD hardware, increasing training contexts by upto 2.7$\times$ and 2.5$\times$ respectively over competitive hand-written baseline at negligible cost to runtime performance.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
CO-EVO: Co-evolving Semantic Anchoring and Style Diversification for Federated DG-ReID
arXiv:2604.26363v1 Announce Type: new
Abstract: Federated domain generalization for person re-identification (FedDG-ReID) aims to collaboratively train a pedestrian retrieval model across multiple decentralized source domains such that it can generalize to unseen target environments without compromising raw data privacy. However, this task is significantly challenged by the inherent stylistic gaps across decentralized clients. Without global supervision, models easily succumb to shortcut learning where representations overfit to domain specific camera biases rather than universal identity features. We propose CO-EVO, a novel federated framework that resolves this semantic-style conflict through a co-evolutionary mechanism. On the semantic side, Camera-Invariant Semantic Anchoring (CSA) learns identity prompts with cross-camera consistency to establish purified and domain-agnostic anchors that filter out local imaging noise. On the visual side, Global Style Diversification (GSD), powered by a Global Camera-Style Bank (GCSB), synthesizes realistic perturbations to expand the visual boundaries of training data. The core of CO-EVO is its co-evolutionary loop where purified anchors act as gravitational centers to guide the image encoder toward robust anatomical attributes amidst diverse style variations. Extensive experiments demonstrate that CO-EVO achieves state-of-the-art (SOTA) performance, proving that the synergy between semantic purification and style expansion is essential for robust cross-domain generalization. Our code is available at: https://github.com/NanYiyuzurn/ACL-LGPS-2026.
Fonte: arXiv cs.CV
RL • Score 85
A Survey of Multi-Agent Deep Reinforcement Learning with Graph Neural Network-Based Communication
arXiv:2604.25972v1 Announce Type: cross
Abstract: In multi-agent reinforcement learning (MARL), the integration of a communication mechanism, allowing agents to better learn to coordinate their actions and converge on their objectives by sharing information. Based on an interaction graph, a subclass of methods employs graph neural networks (GNNs) to learn the communication, enabling agents to improve their internal representations by enriching them with information exchanged. With growing research, we note a lack of explicit structure and framework to distinguish and classify MARL approaches with communication based on GNNs. Thus, this paper surveys recent works in this field. We propose a generalized GNN-based communication process with the goal of making the underlying concepts behind the methods more obvious and accessible.
Fonte: arXiv cs.AI
Privacy/Security/Fairness • Score 85
Privacy-Preserving Federated Learning Framework for Distributed Chemical Process Optimization
arXiv:2604.26073v1 Announce Type: cross
Abstract: Industrial chemical plants often operate under strict data confidentiality constraints, making centralized data-driven process modeling difficult. Federated learning (FL) provides a promising solution by enabling collaborative model training across distributed facilities without sharing raw operational data. This paper proposes a privacy-preserving federated learning framework for distributed chemical process optimization using data collected from multiple geographically separated plants. Each plant locally trains a neural-network-based process model using its own time-series sensor data, while only model parameters are transmitted to a central aggregation server through secure aggregation mechanisms. This design allows cross-plant knowledge sharing while maintaining strict data locality and industrial confidentiality. Experimental evaluation was conducted using process datasets from three independent chemical plants operating under heterogeneous conditions. The results demonstrate rapid convergence of the federated model, with the global mean squared error decreasing from approximately 2369 to below 50 within the first five communication rounds and stabilizing around 35 after 40 rounds. In comparison with local-only training, the proposed federated framework significantly improves prediction accuracy across all plants, while achieving performance comparable to centralized training. The findings indicate that federated learning provides an effective and scalable solution for collaborative industrial analytics, enabling privacy-preserving predictive modeling and process optimization across distributed chemical production facilities.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Adaptive and Fine-grained Module-wise Expert Pruning for Efficient LoRA-MoE Fine-Tuning
arXiv:2604.26340v1 Announce Type: new
Abstract: LoRA-MoE has emerged as an effective paradigm for parameter-efficient fine-tuning, combining the low training cost of LoRA with the increased adaptation capacity of Mixture-of-Experts (MoE). However, existing LoRA-MoE frameworks typically adopt a fixed and uniform expert configuration across heterogeneous Transformer modules (\eg, attention query/key projections and MLP gating networks), ignoring their distinct functional roles and capacity requirements. This design leads to localized over-provisioning, redundant trainable parameters, and unnecessary optimizer-state overhead. Moreover, prior methods enforce load balancing among experts throughout training. Although beneficial in the early stage, this constraint becomes restrictive once routing patterns stabilize, limiting expert specialization on downstream tasks. In this paper, we propose DMEP, a novel LoRA-MoE fine-tuning framework based on Dynamic Module-wise Expert Pruning. DMEP tracks expert utilization during training and physically removes low-utility experts on a per-module basis, yielding a more compact expert structure tailored to different modules. The pruned model then continues training without the load-balancing constraint, freeing the remaining experts to focus entirely on the downstream task and develop specialized expertise. By jointly adapting module-wise expert capacity and eliminating unnecessary balancing, DMEP improves both parameter efficiency and training efficiency. Extensive experiments on multiple reasoning benchmarks show that DMEP reduces trainable parameters by 35\%--43\% and improves training throughput by about 10\%, while maintaining or surpassing the downstream reasoning accuracy of uniform LoRA-MoE baselines.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
AlphaJet: Automated Conceptual Aircraft Synthesis via Disentangled Generative Priors and Topology-Preserving Evolutionary Search
arXiv:2604.26337v1 Announce Type: new
Abstract: Conceptual aircraft design is traditionally an expert-mediated iterative process in which a human designer proposes a configuration, runs low-order physics, inspects the result, and re-proposes. We present AlphaJet, an end-to-end automated synthesis pipeline that closes this loop. From a textual mission specification (mass, range, cruise speed, hard size envelope, engine count, areal density) AlphaJet evolves a feasible 3D aircraft in real time, scored by a transparent multi-disciplinary fitness function covering aerodynamics, structures, weights, stability, packaging, and geometric mount consistency. Three contributions distinguish our approach: (i) an Anatomically-Disentangled Variational Autoencoder (AD-VAE) whose first 25 latent dimensions are supervised to align with named anatomical parameters, providing an interpretable shape prior; (ii) a topology-elitist genetic algorithm that protects the best individual from each of five tail topologies and triggers stagnation restarts, preventing premature collapse to a single configuration; and (iii) mount-aware geometric scoring that computes signed penetration between engines and other structural parts, eliminating the redundant artifacts common in generative aircraft models. The full loop runs interactively on a CPU and streams every generation to a browser viewer, making it a practical real-world automation tool for early-phase design-space exploration.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
OMEGA: Optimizing Machine Learning by Evaluating Generated Algorithms
arXiv:2604.26211v1 Announce Type: new
Abstract: In order to automate AI research we introduce a full, end-to-end framework, OMEGA: Optimizing Machine learning by Evaluating Generated Algorithms, that starts at idea generation and ends with executable code. Our system combines structured meta-prompt engineering with executable code generation to create new ML classifiers. The OMEGA framework has been utilized to generate several novel algorithms that outperform scikit-learn baselines across a robust selection of 20 benchmark datasets (infinity-bench). You can access models discussed in this paper and more in the python package: pip install omega-models.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
AMMA: A Multi-Chiplet Memory-Centric Architecture for Low-Latency 1M Context Attention Serving
arXiv:2604.26103v1 Announce Type: cross
Abstract: All current LLM serving systems place the GPU at the center, from production-level attention-FFN disaggregation to NVIDIA's Rubin GPU-LPU heterogeneous platform. Even academic PIM/PNM proposals still treat the GPU as the central hub for cross-device communication. Yet the GPU's compute-rich architecture is fundamentally mismatched with the memory-bound nature of decode-phase attention, inflating serving latency while wasting power and die area on idle compute units. The problem is compounded as reasoning and agentic workloads push context lengths toward one million tokens, making attention latency the primary user-facing bottleneck.
To address these inefficiencies, we present AMMA, a multi-chiplet, memory-centric architecture for low-latency long-context attention. AMMA replaces GPU compute dies with HBM-PNM cubes, roughly doubling the available memory bandwidth to better serve memory-bound attention workloads. To translate this bandwidth into proportional performance gains, we introduce (i) a logic-die microarchitecture that fully exploits per-cube internal bandwidth for decode attention under a minimal power and area budget, (ii) a two-level hybrid parallelism scheme, and (iii) a reordered collective flow that reduces intra-chip die-to-die communication overhead. We further conduct a design-space exploration over per-cube compute power and intra-chip D2D link bandwidth, providing actionable guidance for hardware designers. Evaluations show that AMMA achieves 15.5X lower attention latency and 6.9X lower energy consumption compared with the NVIDIA H100.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Who Trains Matters: Federated Learning under Enrollment and Participation Selection Biases
arXiv:2604.26604v1 Announce Type: new
Abstract: Federated learning (FL) trains a shared model from updates contributed by distributed clients, often implicitly assuming that contributing clients are representative of the target population. In practice, this representativeness assumption can fail at two distinct stages, inducing selection bias. First, eligibility rules such as device constraints, software requirements, or user consent determine which clients are ever enrolled and reachable for training, inducing \emph{enrollment bias}. Second, among enrolled clients, user and system factors such as battery state, network status, and local time determine which clients participate in each communication round, inducing \emph{participation bias}. Although existing work has largely addressed round-level participation bias, it has paid far less attention to population-level enrollment bias, which can induce a persistent mismatch between the training objective and the target-population objective. We formalize FL under a two-stage selection model and derive \textsc{FedIPW}, an inverse-probability-weighted aggregation scheme that recovers the target-population mean update under standard ignorability and positivity assumptions. Because client-level covariates are often unavailable for non-enrolled clients, we also introduce a limited-information aggregate-calibration extension that uses known target-population summaries to reweight the enrolled sample, partially correcting enrollment bias. We further provide an algorithm-agnostic optimization analysis under residual weighting error and show that incomplete selection correction can induce a non-vanishing bias floor. Finally, experiments on synthetic federated logistic regression validate the predicted objective mismatch and show that enrollment correction reduces target-population error under two-stage selection.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Folding Tensor and Sequence Parallelism for Memory-Efficient Transformer Training & Inference
arXiv:2604.26294v1 Announce Type: new
Abstract: We present tensor and sequence parallelism (TSP), a parallel execution strategy that folds tensor parallelism and sequence parallelism onto a single device axis. In conventional multi-dimensional parallelism layouts, tensor parallelism (TP) shards model weights while sequence parallelism (SP) shards tokens, reducing per-device parameter or activation memory, respectively. Traditionally, each scheme is assigned its own mesh dimension. TSP instead assigns each rank both a weight shard and a sequence shard, reducing both parameter and activation memory along the same device axis. We implement this design with two runtime schedules. For attention, ranks iterate over broadcast parameter shards and reconstruct context through a sequence-wise key/value exchange. For gated MLPs, weight shards circulate in a ring while partial outputs accumulate locally. By sharding both weights and activations across the same devices, TSP trades additional communication volume for reduced memory overhead. We provide a theoretical communication and memory analysis, describe our implementation of TSP attention and gated MLP blocks, and benchmark TSP against TP, SP, and TP+SP. These results position TSP as a hardware-aware alternative for long-context and memory-constrained model training, and as a viable axis of parallelism in concert with existing parallelism schemes such as pipeline and expert parallelism for dense and mixture-of-expert models.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
Federated Medical Image Classification under Class and Domain Imbalance exploiting Synthetic Sample Generation
arXiv:2604.26324v1 Announce Type: new
Abstract: Exploiting deep learning in medical imaging faces critical challenges, including strict privacy constraints, heterogeneous imaging devices with varying acquisition properties, and class imbalance due to the uneven prevalence of pathologies. In this work, we propose FedSSG, a novel Federated Learning framework that addresses domain shifts caused by diverse imaging devices while mitigating the under-representation of rare pathologies. The key contribution is a strategy for generating synthetic samples and distributing them across clients to improve coverage of both underrepresented pathologies and imaging devices. Experimental results demonstrate that our approach significantly enhances model performance and generalization across heterogeneous institutions, with minimal computational overhead at the client side.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
Lightweight Quantum Agent for Edge Systems: Joint PQC and NOMA Resource Allocation
arXiv:2604.25980v1 Announce Type: cross
Abstract: In the context of quantum secure scenarios, existing research on mobile edge devices and intelligent computing and edge (ICE) systems based on the Non-Orthogonal Multiple Access (NOMA) communication model have overlooked the energy consumption overhead of Post-Quantum Cryptography (PQC) modules, and the high complexity of traditional resource allocation algorithms fails to meet the demands of real-time decision-making. To address these challenges, this paper proposes a lightweight agentic AI framework designed for online joint optimization within ICE-enabled mobile devices. The scheme constructs a multi-stage stochastic Mixed Integer Nonlinear Programming (MINLP) model that incorporates static power-consumption constraints for PQC modules. Based on Lyapunov optimization theory, the long-term optimization problem is decoupled, and a linear complexity algorithm is proposed to solve the nonconvex challenges of NOMA power allocation . Simulation results verify that the proposed scheme significantly improves computational throughput while ensuring system queue stability and energy consumption constraints. Compared with traditional Successive Convex Approximation (SCA) algorithms, the complexity is reduced to $\mathcal{O}(N)$, achieving a speedup of approximately 46 times when the number of devices $N=35$, thereby meeting the real-time decision-making requirements in dynamic wireless environments.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Breaking the Autoregressive Chain: Hyper-Parallel Decoding for Efficient LLM-Based Attribute Value Extraction
arXiv:2604.26209v1 Announce Type: new
Abstract: Some text generation tasks, such as Attribute Value Extraction (AVE), require decoding multiple independent sequences from the same document context. While standard autoregressive decoding is slow due to its sequential nature, the independence between output sequences offers an opportunity for parallelism. We present Hyper-Parallel Decoding, a novel decoding algorithm that accelerates offline decoding by leveraging both shared memory and computation across batches. HPD enables out-of-order token generation through position ID manipulation, significantly improving efficiency. Experiments on AVE show that attribute-value pairs are conditionally independent, enabling us to parallelize value generation within each prompt. By further stacking multiple documents within a single prompt, we can decode in parallel up to 96 tokens per prompt. HPD works with all LLMs, and reduces both inference costs and total inference time by up to 13.8X without compromising output quality, potentially saving hundreds of thousands of dollars on industry AVE tasks. Although designed for attribute extraction, HPD makes no assumptions unique to the AVE domain and can in theory be applied to other scenarios with independent output structures.
Fonte: arXiv cs.CL
Theory/Optimization • Score 85
The Serial Scaling Hypothesis
arXiv:2507.12549v4 Announce Type: replace-cross
Abstract: While machine learning has advanced through massive parallelization, we identify a critical blind spot: some problems are fundamentally sequential. These "inherently serial" problems-from mathematical reasoning to physical simulations to sequential decision-making-require sequentially dependent computational steps that cannot be efficiently parallelized. We formalize this distinction in complexity theory, and demonstrate that current parallel-centric architectures face fundamental limitations on such tasks. Then, we show for first time that diffusion models despite their sequential nature are incapable of solving inherently serial problems. We argue that recognizing the serial nature of computation holds profound implications on machine learning, model design, and hardware development.
Fonte: arXiv stat.ML
MLOps/Systems • Score 85
Beyond Fixed Formulas: Data-Driven Linear Predictor for Efficient Diffusion Models
arXiv:2604.26365v1 Announce Type: new
Abstract: To address the high sampling cost of Diffusion Transformers (DiTs), feature caching offers a training-free acceleration method. However, existing methods rely on hand-crafted forecasting formulas that fail under aggressive skipping. We propose L2P (Learnable Linear Predictor), a simple data-driven caching framework that replaces fixed coefficients with learnable per-timestep weights. Rapidly trained in ~20 seconds on a single GPU, L2P accurately reconstructs current features from past trajectories. L2P significantly outperforms existing baselines: it achieves a 4.55x FLOPs reduction and 4.15x latency speedup on FLUX.1-dev, and maintains high visual fidelity under up to 7.18x acceleration on Qwen-Image models, where prior methods show noticeable quality degradation. Our results show learning linear predictors is highly effective for efficient DiT inference. Code is available at https://github.com/Aredstone/L2P-Cache.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
FlowBot: Inducing LLM Workflows with Bilevel Optimization and Textual Gradients
arXiv:2604.26258v1 Announce Type: new
Abstract: LLM workflows, which coordinate structured calls to individual LLMs (each augmented with varying instructions and tools) to achieve a particular goal, offer a promising path towards extending the capabilities of LLMs and building powerful systems that can tackle diverse tasks. However, existing approaches for building such workflows generally rely on human-crafted pipelines and prompts, which presents a substantial bottleneck in real world deployment. How can automatically induce and optimize such workflows in a data-driven way? This paper describes a simple data-driven approach for automatically inducing LLM workflows. We formulate workflow induction as a bilevel optimization problem: an outer loop which optimizes a high-level sketch of the workflow (in particular how the LLM calls should be structured), and an inner loop which optimizes each individual LLM call one-by one. Both loops are optimized with ``textual gradients'' where for the inner loop we optimize each component in a modular way through ``backpropagating'' textual gradients layer-by-layer. We find that LLM workflows discovered through our \textsc{FlowBot} (work\textbf{flow} induction through \textbf{b}ilevel \textbf{o}ptimization and \textbf{t}extual gradients) approach performs competitively against strong baselines that make use of human-crafted or automatically-generated workflows.
Fonte: arXiv cs.CL
MLOps/Systems • Score 85
Unifying Runtime Monitoring Approaches for Safety-Critical Machine Learning: Application to Vision-Based Landing
arXiv:2604.26411v1 Announce Type: new
Abstract: Runtime monitoring is essential to ensure the safety of ML applications in safety-critical domains. However, current research is fragmented, with independent methods emerging from different communities. In this paper, we propose a unified framework categorising runtime monitoring approaches into three distinct types: Operational Design Domain (ODD) monitoring, which ensures compliance with expected operating conditions; Out-of-Distribution (OOD) monitoring, which rejects inputs that deviate from the training data; and Out-of-Model-Scope (OMS) monitoring, which detects anomalous model behaviour based its internal states or outputs. We demonstrate the benefits of this categorization with a dedicated experiment on an aeronautical safety-critical application: runway detection during landing. This framework facilitates design of monitoring activities, with complementary categories of monitors, and enables evaluation and comparison of different monitors using common, safety-oriented metrics.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Co-Learning Port-Hamiltonian Systems and Optimal Energy-Shaping Control
arXiv:2604.26172v1 Announce Type: cross
Abstract: We develop a physics-informed learning framework for energy-shaping control of port-Hamiltonian (pH) systems from trajectory data. The proposed approach {co-learns} a pH system model and an optimal energy-balancing passivity-based controller (EB-PBC) through alternating optimization with policy-aware data collection. At each iteration, the system model is refined using trajectory data collected under the current control policy, and the controller is re-optimized on the updated model. Both components are parameterized by neural networks that embed the pH {dynamics} and EB-PBC structure, ensuring interpretability in terms of energy {interactions}. The learned controller renders the closed-loop system inherently passive and provably stable, and exploits passive plant dynamics without canceling the natural potential. A dissipation regularization enforces strict energy decay during training, thereby enhancing robustness to sim-to-real gaps. The proposed framework is validated on state-regulation and swing-up tasks for planar and torsional pendulum systems.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
Random Cloud: Finding Minimal Neural Architectures Without Training
arXiv:2604.26830v1 Announce Type: new
Abstract: I propose the \emph{Random Cloud} method, a training-free approach to neural architecture search that discovers minimal feedforward network topologies through stochastic exploration and progressive structural reduction. Unlike post-training pruning methods that require a full train-prune-retrain cycle, this method evaluates randomly initialized networks without backpropagation, progressively reduces their topology, and only trains the best minimal candidate at the end. I evaluate on 7 classification benchmarks against magnitude pruning and random pruning baselines. The Random Cloud matches or outperforms both baselines in 6 of 7 datasets, achieving statistically significant improvements on Sonar ($+4.9$pp accuracy, $p{=}0.017$ vs magnitude pruning) with 87\% parameter reduction. Crucially, the method is faster than both pruning baselines in 4 of 5 datasets (0.67--0.94$\times$ the cost of full training), since it avoids training the full-size network entirely.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Operating-Layer Controls for Onchain Language-Model Agents Under Real Capital
arXiv:2604.26091v1 Announce Type: new
Abstract: We study reliability in autonomous language-model agents that translate user mandates into validated tool actions under real capital. The setting is DX Terminal Pro, a 21-day deployment in which 3,505 user-funded agents traded real ETH in a bounded onchain market. Users configured vaults through structured controls and natural-language strategies, but only agents could choose normal buy/sell trades. The system produced 7.5M agent invocations, roughly 300K onchain actions, about $20M in volume, more than 5,000 ETH deployed, roughly 70B inference tokens, and 99.9% settlement success for policy-valid submitted transactions. Long-running agents accumulated thousands of sequential decisions, including 6,000+ prompt-state-action cycles for continuously active agents, yielding a large-scale trace from user mandate to rendered prompt, reasoning, validation, portfolio state, and settlement. Reliability did not come from the base model alone; it emerged from the operating layer around the model: prompt compilation, typed controls, policy validation, execution guards, memory design, and trace-level observability. Pre-launch testing exposed failures that text-only benchmarks rarely measure, including fabricated trading rules, fee paralysis, numeric anchoring, cadence trading, and misread tokenomics. Targeted harness changes reduced fabricated sell rules from 57% to 3%, reduced fee-led observations from 32.5% to below 10%, and increased capital deployment from 42.9% to 78.0% in an affected test population. We show that capital-managing agents should be evaluated across the full path from user mandate to prompt, validated action, and settlement.
Fonte: arXiv cs.AI
Privacy/Security/Fairness • Score 85
DP-CDA: An Algorithm for Enhanced Privacy Preservation in Dataset Synthesis Through Randomized Mixing
arXiv:2411.16121v3 Announce Type: replace
Abstract: In recent years, the growth of data across various sectors, including healthcare, security, finance, and education, has created significant opportunities for analysis and informed decision-making. However, these datasets often contain sensitive and personal information, which raises serious privacy concerns. It has been shown in multiple works that a person's identity is intertwined with their data, even if the data is anonymized. Due to this lack of separation between a person's identity and their information, the patterns associated with an individual's information can uniquely identify them. Protecting individual privacy is crucial, yet many existing machine learning and data publishing algorithms struggle with high-dimensional data, facing challenges related to the trade-off between computational efficiency and privacy. To address these challenges, we introduce an effective data publishing algorithm \emph{DP-CDA}. Our proposed algorithm generates synthetic data by randomly mixing the privacy-sensitive data in a class-specific manner and inducing carefully tuned randomness to ensure formal privacy guarantees. Our comprehensive privacy accounting shows that the proposed DP-CDA provides a stronger privacy guarantee compared to existing methods, allowing for better utility while maintaining a stricter level of privacy. To evaluate the effectiveness of DP-CDA, we examine the accuracy of predictive models trained on the synthetic data, which serves as a measure of dataset utility. Importantly, we identify an optimal order of mixing that balances privacy-utility trade-off. Our results indicate that synthetic datasets produced using the DP-CDA can achieve superior utility compared to those generated by conventional data publishing algorithms, even when subject to the same privacy requirements.
Fonte: arXiv stat.ML
Theory/Optimization • Score 85
Latent Autoencoder Ensemble Kalman Filter for Nonlinear Data assimilation
arXiv:2603.06752v2 Announce Type: replace-cross
Abstract: The ensemble Kalman filter (EnKF) is widely used for data assimilation in high-dimensional systems, but its performance often deteriorates for strongly nonlinear dynamics due to the structural mismatch between the Kalman update and the underlying system behavior. In this work, we propose a latent autoencoder ensemble Kalman filter (LAE-EnKF) that addresses this limitation by reformulating the assimilation problem in a learned latent space with linear and stable dynamics. The proposed method learns a nonlinear encoder--decoder together with a stable linear latent evolution operator and a consistent latent observation mapping, yielding a closed linear state-space model in the latent coordinates. This construction restores compatibility with the Kalman filtering framework and allows both forecast and analysis steps to be carried out entirely in the latent space. Compared with existing autoencoder-based and latent assimilation approaches that rely on unconstrained nonlinear latent dynamics, the proposed formulation emphasizes structural consistency, stability, and interpretability. We provide a theoretical analysis of learning linear dynamics on low-dimensional manifolds and establish generalization error bounds for the proposed latent model. Numerical experiments on representative nonlinear and chaotic systems demonstrate that the LAE-EnKF yields more accurate and stable assimilation than the standard EnKF and related latent-space methods, while maintaining comparable computational cost and data-driven.
Fonte: arXiv stat.ML
RL • Score 90
DORA: A Scalable Asynchronous Reinforcement Learning System for Language Model Training
arXiv:2604.26256v1 Announce Type: new
Abstract: Reinforcement learning (RL) has become a critical paradigm for LLM post-training, yet the rollout phase -- accounting for 50--80% of total step time -- is bottlenecked by skewed generation: long-tailed trajectories indispensable for model performance block the entire training pipeline. Asynchronous training offers a natural remedy by overlapping generation with training, but introduces a fundamental tension between efficiency and algorithmic correctness. We identify three constraints in asynchronous training to preserve convergence: intra-trajectory policy consistency, data integrity, and bounded staleness. Existing approaches fail to intrinsically address the long-tailed trajectory problem, which is further exacerbated by the imbalance characteristic of Mix-of-Experts models, or deviate from the standard RL training formulation, thereby hindering model convergence. Therefore, we propose DORA (Dynamic ORchestration for Asynchronous Rollout), which addresses this challenge through algorithm-system co-design. DORA introduces multi-version streaming rollout, a novel asynchronous paradigm that maintains multiple policy versions concurrently -- simultaneously achieving full bubble elimination without compromising algorithmic constraints. Experimental results demonstrate that our DORA system achieves substantial improvements in throughput -- up to 2--3 times higher than state-of-the-art systems on open-source benchmarks -- without compromising convergence. Furthermore, in large-scale industrial applications with tens of thousands of accelerators, DORA accelerates RL training by 2--4 times compared to synchronous training across various scenarios. The resultant open-source models, LongCat-Flash-Thinking, exhibit competitive performance on complex reasoning benchmarks, matching the capability of most advanced LLMs.
Fonte: arXiv cs.LG
Privacy/Security/Fairness • Score 85
Efficient and Interpretable Transformer for Counterfactual Fairness
arXiv:2604.26188v1 Announce Type: new
Abstract: The growing reliance of machine learning models in high-stakes, highly regulated domains such as finance and insurance has created a growing tension between predictive performance, interpretability, and regulatory fairness requirements. In these settings, models are expected not only to deliver reliable predictions but also to provide transparent decision rationales and comply with strict fairness requirements. Attention-based transformers offer powerful mechanisms for modeling complex data relationships as demonstrated in various language tasks, yet their attention mechanisms alone do not ensure counterfactually fair predictions, even when combined with fairness-aware techniques. To address these limitations, we propose the Feature Correlation Transformer (FCorrTransformer), an attention-light architecture tailored for tabular data. In this design, the attention matrix admits a direct statistical interpretation as pairwise feature dependencies, enhancing both interpretability and efficiency. Leveraging this structure, we introduce Counterfactual Attention Regularization (CAR), a framework that enforces group-invariant fair representations of sensitive features at the attention level, promoting counterfactually fair predictions without relying on explicit causal assumptions. Empirical evaluations on imbalanced classification and regression benchmarks demonstrate that FCorrTransformer combined with CAR achieves strong counterfactual fairness while maintaining competitive predictive performance and substantially reducing model complexity compared with standard transformer-based baselines. Overall, this work bridges a critical gap between fairness theory and machine learning models, offering a practical framework for responsible AI in regulatory-sensitive domains.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
STLGT: A Scalable Trace-Based Linear Graph Transformer for Tail Latency Prediction in Microservices
arXiv:2604.26422v1 Announce Type: new
Abstract: Accurate end-to-end tail-latency forecasting is critical for proactive SLO management in microservice systems. However, modeling long-range dependency propagation and non-stationary, bursty workloads while maintaining inference efficiency at scale remains challenging. We present STLGT (Scalable Trace-based Linear Graph Transformer), a per-API predictor that encodes traces as span graphs for multi-step p95 tail-latency forecasting. STLGT uses a structure-aware linear graph Transformer to propagate cross-service dependencies with inference time linear in span graph size, and a decoupled temporal module to capture workload dynamics. Across a personalized education microservice application, DeathStarBench, and Alibaba traces, STLGT improves forecasting accuracy over PERT-GNN by 8.5% MAPE on average and achieves up to 12x faster CPU inference at N=32, matching the maximum span graph size after preprocessing the Alibaba traces. Ablation studies further demonstrate the effectiveness of each component, especially under bursty traffic.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
CoQuant: Joint Weight-Activation Subspace Projection for Mixed-Precision LLMs
arXiv:2604.26378v1 Announce Type: new
Abstract: Post-training quantization (PTQ) has become an important technique for reducing the inference cost of Large Language Models (LLMs). While recent mixed-precision methods improve ultra-low bit quantization by preserving critical subspaces in high precision, they typically construct these subspaces relying solely on activation statistics. This ignores the fundamental nature of linear operations, where the output perturbation is jointly driven by both activation and weight quantization noise. In this paper, we propose CoQuant, a joint weight-activation subspace projection method. By theoretically modeling the expected output error, CoQuant formulates a closed-form weighted PCA solution that balances activation and weight covariances to select the optimal high-precision subspace. Extensive experiments on Llama-3.2 and Qwen2.5 models show that CoQuant consistently outperforms strong PTQ baselines in both WikiText perplexity and zero-shot common-sense reasoning accuracy. These results demonstrate that joint weight-activation subspace modeling provides a principled and effective direction for low-bit LLM quantization. The source code is available at https://github.com/Zachary5895/CoQuant.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
RaMP: Runtime-Aware Megakernel Polymorphism for Mixture-of-Experts
arXiv:2604.26039v1 Announce Type: cross
Abstract: The optimal kernel configuration for Mixture-of-Experts (MoE) inference depends on both batch size and the expert routing distribution, yet production systems dispatch from batch size alone, leaving 10-70% of kernel throughput unrealized. We present RaMP, a routing-aware dispatch framework. A performance-region analysis derives, from hardware constants alone, when each optimization helps, correctly predicting all 8 tested architectures, including 3 unseen. A four-parameter wave cost model selects the fastest configuration from the runtime expert histogram, achieving 0.93% mean regret versus exhaustive search, fitted from just 10-24 minutes of one-time profiling per model. Because the model depends only on CTA grid geometry, it is kernel-agnostic: applied to Alpha-MoE, it delivers 1.14x with no source modification. Paired with a co-designed CuTe DSL kernel exposing 134-268 polymorphic configurations, RaMP delivers 1.22x kernel speedup over static dispatch and 1.30x end-to-end speedup in vLLM serving over Triton, 1.41x over DeepGEMM, and 1.13x over FlashInfer CUTLASS.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Rethinking KV Cache Eviction via a Unified Information-Theoretic Objective
arXiv:2604.25975v1 Announce Type: cross
Abstract: Key-value (KV) caching is essential for large language model inference, yet its memory overhead poses a critical bottleneck for long-context generation. Existing eviction policies predominantly rely on empirical heuristics, lacking a rigorous theoretical foundation. This work rethinks KV cache eviction through the lens of the Information Bottleneck principle. Under a linear-Gaussian surrogate of attention, we derive a closed-form mutual information objective that characterizes the effective information capacity of a retained KV cache subset. This formulation reveals that a wide range of existing eviction strategies can be interpreted as different approximations of the same capacity-maximization principle. Guided by this insight, we introduce CapKV, a capacity-aware eviction method that directly targets information preservation via a log-determinant approximation using statistical leverage scores. This approach replaces heuristic selection with a theoretically grounded mechanism that preserves the maximum predictive signal. Extensive experiments across multiple models and long-context benchmarks show that CapKV consistently outperforms prior methods, achieving a better trade-off between memory efficiency and generational fidelity.
Fonte: arXiv cs.AI
RL • Score 85
Accelerating RL Post-Training Rollouts via System-Integrated Speculative Decoding
arXiv:2604.26779v1 Announce Type: new
Abstract: RL post-training of frontier language models is increasingly bottlenecked by autoregressive rollout generation, making rollout acceleration a central systems challenge. Many existing efficiency methods improve throughput by changing the rollout or optimization regime, for example, through off-policy execution, replay, or lower-precision generation. We study speculative decoding as a lossless acceleration primitive for RL rollouts that preserves the target model's output distribution. We implement speculative decoding in NeMo-RL with a vLLM backend, supporting both synchronous and asynchronous pipelines and enabling speculation during RL rollouts. This benefit is realizable across speculation mechanisms, such as pretrained MTP heads, small external draft models or even techniques such as Eagle3, which are traditionally applied after RL phase. This yields a deployment path for state-of-the-art speculative decoding inside RL training. In a reasoning post-training workload at 8B scale under synchronous RL, speculative decoding improves rollout throughput by 1.8x. Using a high-fidelity performance simulator, we project that combining speculative decoding with asynchronous RL yields up to 2.5x end-to-end training speedup at 235B scale.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Hierarchical adaptive control for real-time dynamic inference at the edge
arXiv:2604.26470v1 Announce Type: new
Abstract: Industrial systems increasingly depend on Machine Learning (ML), and operate on heterogeneous nodes that must satisfy tight latency, energy, and memory constraints. Dynamic ML models, which reconfigure their computational footprint at runtime, promise high energy efficiency and lower average latency for modest accuracy tradeoffs; however, their deployment is complex due to the additional hyperparameters they rely on. These hyperparameters, controlling the accuracy versus average latency tradeoff, are often tuned on a calibration dataset that must match the test time distribution, an assumption that rarely holds in real-world scenarios, leading to suboptimal operational conditions, possibly below static models. We propose a two-tier adaptive architecture that co-optimizes model and system decisions. At the global level, a scheduler configures and deploys, for each edge node, a cascade of classifiers composed of lightweight specialized models and a generalist fallback, satisfying latency and memory constraints. At the node level, a local controller tracks data drifts and hardware resources, enabling or disabling specialized predictors (SP) to preserve high energy efficiency and avoid latency-constraint violations under varying conditions. This design allows longer operating times without forcing a global redeployment step, and enables efficient execution in case of an unreachable remote global controller. We evaluate the approach on two datasets under controlled distribution mismatch scenarios, showing average per-inference reductions of latency up to 2.45x and energy up to 2.86x, with less than 4% accuracy drop compared to static baselines. Our contributions are:(1) a budgeted SP-cascade formulation that preserves worst-case latency constraints;(2) a hierarchical controller that maintains efficiency under data and resource changes; and (3) an experimental evaluation on embedded hardware.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
SciHorizon-DataEVA: An Agentic System for AI-Readiness Evaluation of Heterogeneous Scientific Data
arXiv:2604.26645v1 Announce Type: new
Abstract: AI-for-Science (AI4Science) is increasingly transforming scientific discovery by embedding machine learning models into prediction, simulation, and hypothesis generation workflows across domains. However, the effectiveness of these models is fundamentally constrained by the AI-readiness of scientific data, for which no scalable and systematic evaluation mechanism currently exists. In this work, we propose SciHorizon-DataEVA, a novel agentic system to scalable AI-readiness evaluation of heterogeneous scientific data. At the evaluation-criteria level, we introduce the Sci-TQA2 principles, which organize AI-readiness into four complementary dimensions: Governance Trustworthiness, Data Quality, AI Compatibility, and Scientific Adaptability. Each dimension is decomposed into measurable atomic elements that enable fine-grained and executable assessment. To operationalize these principles at scale, we develop Sci-TQA2-Eval, a hierarchical multi-agent evaluation approach orchestrated through a directed, cyclic workflow. Our Sci-TQA2-Eval dynamically constructs dataset-aware evaluation specifications by combining lightweight dataset profiling, applicability-aware metric activation, and knowledge-augmented planning grounded in domain constraints and dataset-paper signals. These specifications are executed through an adaptive, tool-centric evaluation mechanism with built-in verification and self-correction, enabling scalable and reliable assessment across heterogeneous scientific data. Extensive experiments on scientific datasets spanning multiple domains demonstrate the effectiveness and generality of SciHorizon-DataEVA for principled AI-readiness evaluation.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
Super-resolution Multi-signal Direction-of-Arrival Estimation by Hankel-structured Sensing and Decomposition
arXiv:2604.26793v1 Announce Type: new
Abstract: Motivated by sensing modalities in modern autonomous systems that involve hardware-constrained spatial sampling over large arrays with limited coherence time, we develop a novel framework for rapid super-resolution multi-signal direction-of-arrival (DoA) estimation based on Hankel-structured sensing and data matrix decomposition of arbitrary rank, under both the $L_2$ and $L_1$-norm formulation. The resulting $L_2$-norm estimator is shown to be maximum-likelihood optimal in white Gaussian noise. The $L_1$-norm estimator is shown to be maximum-likelihood optimal in independent, identically distributed (i.i.d.) isotropic Laplace noise, offering broad robustness to impulsive interference and corrupted measurements commonly encountered in practice. Extensive simulations demonstrate that the proposed methods exhibit powerful super-resolution capabilities, requiring significantly lower SNR and achieving substantially higher resolution probability than recent competing approaches.
Fonte: arXiv cs.LG
RL • Score 85
Lifting Embodied World Models for Planning and Control
arXiv:2604.26182v1 Announce Type: new
Abstract: World models of embodied agents predict future observations conditioned on an action taken by the agent. For complex embodiments, action spaces are high-dimensional and difficult to specify: for example, precisely controlling a human agent requires specifying the motion of each joint. This makes the world model hard to control and expensive to plan with as search-based methods like CEM scale poorly with action dimensionality. To address this issue, we train a lightweight policy that maps high-level actions to sequences of low-level joint actions. Composing this policy with the frozen world model produces a lifted world model that predicts a sequence of future observations from a single high-level action. We instantiate this framework for a human-like embodiment, defining the high-level action space as a small set of 2D waypoints annotated on the current observation frame, each specifying a near-term goal position for a leaf joint (pelvis, head, hands). Waypoints are low-dimensional, visually interpretable, and easy to specify manually or to search over. We show that the lifted world model substantially outperforms searching directly in low-level joint space ($3.8\times$ lower mean joint error to the goal pose), while remaining more compute-efficient and generalizing to environments unseen by the policy.
Fonte: arXiv cs.CV
Vision • Score 85
Progressive Semantic Communication for Efficient Edge-Cloud Vision-Language Models
arXiv:2604.26508v1 Announce Type: new
Abstract: Deploying Vision-Language Models (VLMs) on edge devices remains challenging due to their substantial computational and memory demands, which exceed the capabilities of resource-constrained embedded platforms. Conversely, fully offloading inference to the cloud is often impractical in bandwidth-limited environments, where transmitting raw visual data introduces substantial latency overhead. While recent edge-cloud collaborative architectures attempt to partition VLM workloads across devices, they typically rely on transmitting fixed-size representations, lacking adaptability to dynamic network conditions and failing to fully exploit semantic redundancy. In this paper, we propose a progressive semantic communication framework for edge-cloud VLM inference, using a Meta AutoEncoder that compresses visual tokens into adaptive, progressively refinable representations, enabling plug-and-play deployment with off-the-shelf VLMs without additional fine-tuning. This design allows flexible transmission at different information levels, providing a controllable trade-off between communication cost and semantic fidelity. We implement a full end-to-end edge-cloud system comprising an embedded NXP i.MX95 platform and a GPU server, communicating over bandwidth-constrained networks. Experimental results show that, at 1 Mbps uplink, the proposed progressive scheme significantly reduces network latency compared to full-edge and full-cloud solutions, while maintaining high semantic consistency even under high compression. The implementation code will be released upon publication at https://github.com/open-ep/ProSemComVLM.
Fonte: arXiv cs.LG
Multimodal • Score 85
SWAN: World-Aware Adaptive Multimodal Networks for Runtime Variations
arXiv:2604.26181v1 Announce Type: new
Abstract: Multimodal deep neural networks deployed in realistic environments must contend with runtime variations: changes in modality quality, overall input complexity, and available platform resources. Current networks struggle with such fluctuations -- adaptive networks cannot adhere to a strict compute budget, controller-based networks neglect to consider input complexity, and statically provisioned networks fail at all the above. Consequently, they do not extract maximum utility from the expended computational resources. We present SWAN (Sample and World-Aware Multimodal Network), the first adaptive multimodal network that accomplishes all three goals. SWAN employs a quality-aware controller to assign resources among modalities according to a variable user-specified maximum budget. Within this budget, an adaptive gating module further optimizes efficiency by scaling layer utilization according to sample complexity. For further gains, SWAN also employs a token dropping module that masks semantically irrelevant multimodal features before performing detections. We evaluate SWAN in the domain of autonomous driving with complex multi-object 3D detection, reducing FLOPs by up to 49% with minimal degradation.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Bian Que: An Agentic Framework with Flexible Skill Arrangement for Online System Operations
arXiv:2604.26805v1 Announce Type: new
Abstract: Operating and maintaining (O&M) large-scale online engine systems (search, recommendation, advertising) demands substantial human effort for release monitoring, alert response, and root cause analysis. While LLM-based agents are a natural fit for these tasks, the deployment bottleneck is not reasoning capability but orchestration: selecting, for each operational event, the relevant data (metrics, logs, change events) and the applicable operational knowledge (handbook rules and practitioner experience). Feeding all signals indiscriminately causes dilution and hallucination, while manually curating the event-to-(data, knowledge) mapping is intractable under dozens of daily releases. We present Bian Que, an agentic framework with three contributions: (i) a \emph{unified operational paradigm} abstracting day-to-day O&M into three canonical patterns: release interception, proactive inspection, and alert root cause analysis; (ii) \emph{Flexible Skill Arrangement}, where each Skill specifies which data and knowledge to retrieve for a given business-module context and can be automatically generated and updated by LLMs or iteratively refined through natural-language instructions from on-call engineers; (iii) a \emph{unified self-evolving mechanism} in which one correction signal drives two parallel pathways, case-memory-to-knowledge distillation and targeted Skill refinement. Deployed on the e-commerce search engine of KuaiShou, the major short-video platform in China, Bian Que reduces alert volume by 75%, achieves 80% root-cause analysis accuracy, and cuts mean time to resolution by over 50%. Our framework achieves 99.0% pass rate on offline evaluations. Our code is available at https://github.com/benchen4395/BianQue_Assistant.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
Safe, Scalable, and Accurate Bayes Posterior Sampling for Large-Data Generalized Linear Mixed Models
arXiv:2604.26029v1 Announce Type: cross
Abstract: We consider the problem of scalable sampling algorithms to fit Bayesian generalized linear mixed models on large datasets. Stochastic gradient Langevin dynamics, coupled with smooth re-parameterizations of variance parameters, produces divergent Markov chains and cannot be reliably used for sampling covariance parameters of random effects. We advocate the use of a mirror Langevin dynamics algorithm, propose the novel stochastic mirror Langevin dynamics based on data subsampling, and provide concrete guidelines for its use in a Bayesian inference framework. Based on an explicit Wasserstein distance error bound between the posterior and its algorithmic approximation, we propose a post-processing step that yields an asymptotic, order-wise correct estimation of the posterior variance, eliminating the irreducible posterior variance estimation bias due to subsampling. Empirical performance of the method is evaluated through simulated experiments and a longitudinal study of pain trajectories in a study of breast cancer survivors.
Fonte: arXiv stat.ML
Privacy/Security/Fairness • Score 85
Asynchronous Federated Unlearning with Invariance Calibration for Medical Imaging
arXiv:2604.26809v1 Announce Type: new
Abstract: Federated Unlearning (FU) is an emerging paradigm in Federated Learning (FL) that enables participating clients to fully remove their contributions from a trained global model, driven by data protection regulations that mandate the right to be forgotten. However, existing FU methods mostly rely on synchronous coordination. This requirement forces the entire federation to halt and wait for stragglers to complete erasure, creating significant delays due to device heterogeneity. Furthermore, these methods often face the problem that the influence of erased data is merely suppressed temporarily and resurfaces during subsequent training, rather than being genuinely removed. To overcome these limitations, this paper proposes Asynchronous Federated Unlearning with Invariance Calibration (AFU-IC), a novel framework for medical imaging that decouples the erasure process from the global training workflow. This enables the target client to perform unlearning asynchronously without interrupting global training. Meanwhile, a server-side invariance calibration mechanism prevents the model from relearning the erased data. Extensive experiments on three medical benchmarks demonstrate that AFU-IC achieves unlearning efficacy and model fidelity comparable to gold-standard retraining while significantly reducing wall-clock latency compared to synchronous baselines. AFU-IC ensures efficient, compliant and reliable FL in cross-silo medical environments.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Sample Selection Using Multi-Task Autoencoders in Federated Learning with Non-IID Data
arXiv:2604.26116v1 Announce Type: new
Abstract: Federated learning is a machine learning paradigm in which multiple devices collaboratively train a model under the supervision of a central server while ensuring data privacy. However, its performance is often hindered by redundant, malicious, or abnormal samples, leading to model degradation and inefficiency. To overcome these issues, we propose novel sample selection methods for image classification, employing a multitask autoencoder to estimate sample contributions through loss and feature analysis. Our approach incorporates unsupervised outlier detection, using one-class support vector machine (OCSVM), isolation forest (IF), and adaptive loss threshold (AT) methods managed by a central server to filter noisy samples on clients. We also propose a multi-class deep support vector data description (SVDD) loss controlled by a central server to enhance feature-based sample selection. We validate our methods on CIFAR10 and MNIST datasets across varying numbers of clients, non-IID distributions, and noise levels up to 40%. The results show significant accuracy improvements with loss-based sample selection, achieving gains of up to 7.02% on CIFAR10 with OCSVM and 1.83% on MNIST with AT. Additionally, our federated SVDD loss further improves feature-based sample selection, yielding accuracy gains of up to 0.99% on CIFAR10 with OCSVM. These results show the effectiveness of our methods in improving model accuracy across various client counts and noise conditions.
Fonte: arXiv cs.CV
Theory/Optimization • Score 85
A Tutorial Review of Bayesian Optimization with Gaussian Processes to Accelerate Stationary Point Searches
arXiv:2603.10992v5 Announce Type: replace
Abstract: Building local surrogates to accelerate stationary point searches on potential energy surfaces spans decades of effort. Done correctly, surrogates can reduce the number of expensive electronic structure evaluations by roughly an order of magnitude while preserving the accuracy of the underlying theory, with the gain depending on oracle cost, search distance, and the availability of analytical forces. We present a unified Bayesian optimization view of minimization, single-point saddle searches, and double-ended path searches: all three share one six-step surrogate loop and differ only in the inner optimization target and the acquisition criterion. The framework uses Gaussian process regression with derivative observations, inverse-distance kernels, and active learning, and we develop optional extensions for production use, including farthest-point sampling with the Earth Mover's Distance, MAP regularization, an adaptive trust radius, and random Fourier features for scaling. Accompanying pedagogical Rust code demonstrates that all three applications use the same Bayesian optimization loop, bridging the gap between theoretical formulation and practical execution.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
PiGGO: Physics-Guided Learnable Graph Kalman Filters for Virtual Sensing of Nonlinear Dynamic Structures under Uncertainty
arXiv:2604.26593v1 Announce Type: new
Abstract: Digital twins provide a powerful paradigm for diagnostic and prognostic tasks in the monitoring and control of engineered systems; however, their deployment for complex structures remains challenged by model-form uncertainty, arising from unknown nonlinear dynamics, and by sparse sensing. These limitations hinder reliable online state estimation using either purely physics-based or purely data-driven approaches. This work introduces the Physics-Guided Graph Neural ODE (PiGGO) framework, a physics-informed, graph-based Bayesian state estimation approach in which a learned graph neural ordinary differential equation (GNODE) serves as the continuous-time state-transition model within an extended Kalman filter. The graph representation explicitly defines the system state-space, while physics-guided inductive biases encode known structural relationships and constrain the learning of nonlinear dynamics. By integrating graph-native learned dynamics with recursive Bayesian filtering, the proposed PiGGO framework enables online virtual sensing and uncertainty-aware state estimation for nonlinear systems with unknown model form, while maintaining generalisation across topologically similar structures. Numerical case studies demonstrate improved robustness to model uncertainty and measurement noise, outperforming both open-loop graph neural models and conventional filtering approaches in online prediction tasks.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Large-scale semi-supervised learning with online spectral graph sparsification
arXiv:2604.26550v1 Announce Type: new
Abstract: We introduce Sparse-HFS, a scalable algorithm that can compute solutions to SSL problems using only O(n polylog(n)) space and O(m polylog(n)) time.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Accurate and Robust Generative Approach for Overcoming Data Sparsity and Imbalance in Landslide Modeling with A Tabular Foundation Model
arXiv:2604.25159v1 Announce Type: new
Abstract: Landslide investigation relies on sufficient and well-balanced observational data influenced by geological, hydrological, and anthropogenic factors. Available landslide inventories are often sparse and imbalanced, which limits understanding of triggering conditions and failure mechanisms. Data generation provides an effective approach to help capture feature dependencies from limited landslide observations. However, existing generation approaches for landslides often struggle to capture complex relationships among features and lack robustness across multiple scenarios and interacting factors. Here, we propose an accurate and robust approach for generating multi-feature landslide datasets by utilizing a tabular foundation model. By leveraging the capacity to learn from limited observations, the proposed approach effectively preserves the multivariate dependencies and statistical characteristics inherent in landslide occurrences. Comparative experiments on 20 landslide inventories demonstrate that the generated datasets closely align with observed distributions, maintain realistic feature dependencies, and exhibit robustness across different environmental contexts. This work provides an effective approach to overcome data sparsity and imbalance and strengthens landslide susceptibility modeling and risk assessment under limited observations.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Knowledge Distillation Must Account for What It Loses
arXiv:2604.25110v1 Announce Type: new
Abstract: This position paper argues that knowledge distillation must account for what it loses: student models should be judged not only by retained task scores, but by whether they preserve the teacher capabilities that make those scores reliable. This matters because distillation is increasingly used to turn large, often frontier models into deployable systems, yet headline metrics can hide losses in uncertainty, boundary behavior, process reliability, on-policy stability, grounding, privacy, safety, and diversity. We identify the retention assumption behind current evaluation and reframe distillation as a lossy projection of teacher behavior rather than a faithful copy. We then synthesize existing evidence into a taxonomy of off-metric distillation losses, showing that these losses are concrete, recurring, and measurable. To make the position actionable, we propose scenario-specific preservation targets and a Distillation Loss Statement that reports what was preserved, what was lost, and why the remaining losses are acceptable. The goal is not lossless distillation, but accountable distillation.
Fonte: arXiv cs.LG
NLP/LLMs • Score 92
Marco-MoE: Open Multilingual Mixture-of-Expert Language Models with Efficient Upcycling
arXiv:2604.25578v1 Announce Type: new
Abstract: We present Marco-MoE, a suite of fully open multilingual sparse Mixture-of-Experts (MoE) models. Marco-MoE features a highly sparse design in which only around 5\% of the total parameters are activated per input token. This extreme sparsity, combined with upcycling from dense models, enables efficient pre-training on 5T tokens. Our models surpass similarly-sized competitors on English and multilingual benchmarks, achieving a best-in-class performance-to-compute ratio. We further post-train these models to create Marco-MoE-\textsc{Instruct} variants, which surpass the performance of competing models possessing $3$--$14\times$ more activated parameters. Our analysis reveals that Marco-MoE learns structured expert activation patterns shared across related languages, while maintaining highly specialized utilization for linguistically isolated ones. We further show that Marco-MoE allows for scalable language expansion without the interference typical of dense models. To support the community, we disclose our full training datasets, recipes, and model weights.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
BenchGuard: Who Guards the Benchmarks? Automated Auditing of LLM Agent Benchmarks
arXiv:2604.24955v1 Announce Type: new
Abstract: As benchmarks grow in complexity, many apparent agent failures are not failures of the agent at all - they are failures of the benchmark itself: broken specifications, implicit assumptions, and rigid evaluation scripts that penalize valid alternative approaches. We propose employing frontier LLMs as systematic auditors of evaluation infrastructure, and realize this vision through BenchGuard, the first automated auditing framework for task-oriented, execution-based agent benchmarks. BenchGuard cross-verifies all benchmark artifacts via structured LLM protocols, optionally incorporating agent solutions or execution traces as additional diagnostic evidence. Deployed on two prominent scientific benchmarks, BenchGuard identified 12 author-confirmed issues in ScienceAgentBench - including fatal errors rendering tasks unsolvable - and exactly matched 83.3% of expert-identified issues on the BIXBench Verified-50 subset, catching defects that prior human review missed entirely. A full audit of 50 complex bioinformatics tasks costs under USD 15, making automated benchmark auditing a practical and valuable complement to human review. These findings point toward AI-assisted benchmark development, where frontier models serve not only as subjects of evaluation but as active participants in validating the evaluation infrastructure itself.
Fonte: arXiv cs.CL
RL • Score 85
DGLight: DQN-Guided GRPO Fine-Tuning of Large Language Models for Traffic Signal Control
arXiv:2604.25259v1 Announce Type: new
Abstract: Traffic signal control (TSC) plays a central role in reducing congestion and maintaining urban mobility. This dissertation introduces DGLight, a critic-guided reinforcement-learning framework for adapting a pretrained large language model to TSC. DGLight first trains a CoLight-based Deep Q-Network critic to estimate traffic-aware action values from structured intersection states, then uses the frozen critic to score candidate language-model actions and optimize the policy with Group Relative Policy Optimization (GRPO). The resulting controller maps traffic states to interpretable reasoning traces and signal decisions while learning from dense per-state supervision rather than raw cumulative environment rewards. Experiments on TSC benchmarks covering Jinan and Hangzhou show that DGLight is the strongest overall method among the compared LLM-based controllers, remains competitive with strong RL baselines, and transfers well to city datasets not used to fit the critic. Qualitative examples further show that the model's generated reasoning is interpretable and aligned with the chosen signal phase. The project code is available $\href{https://github.com/yyccbb/FYP_LLMTSC}{here}$.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Feasible-First Exploration for Constrained ML Deployment Optimization in Crash-Prone Hierarchical Search Spaces
arXiv:2604.25073v1 Announce Type: new
Abstract: Deploying machine learning models under production constraints requires joint optimization over model family, quantization scheme, runtime backend, and serving configuration. This induces a hierarchical mixed-variable search space in which many configurations are invalid: evaluations may crash, exceed memory limits, or violate latency constraints. Standard black-box optimizers such as Tree-structured Parzen Estimators (TPE) and constrained Bayesian optimization are effective when valid configurations are common, but they can spend a large fraction of a small evaluation budget on invalid or uninformative trials in hostile deployment spaces. This paper studies that regime and asks whether optimization should be decomposed into an explicit exploration stage followed by model-guided exploitation. We propose Thermal Budget Annealing (TBA), a feasible-first exploration procedure that maps valid and feasible regions before warm-starting TPE. The method includes two robustness mechanisms for hostile hardware: trial timeouts that abort clearly infeasible evaluations early, and subspace blacklisting that temporarily suppresses categorical subspaces after repeated failures. We also introduce DeployBench, a benchmark suite for deployment optimization with hierarchical structure, hidden crash zones, hard constraints, and unequal evaluation costs. On synthetic benchmarks and real GPU deployment with five pre-trained vision models across five GPU targets (NVIDIA H100, A100, RTX 5080, L4, and T4), the proposed hybrid improves model-family discovery under tight constraints while reducing wasted budget relative to cold-start TPE.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
A Comparative Analysis on the Performance of Upper Confidence Bound Algorithms in Adaptive Deep Neural Networks
arXiv:2604.24810v1 Announce Type: new
Abstract: Edge computing environments impose strict constraints on energy consumption and latency, making the deployment of deep neural networks a significant challenge. Therefore, smart and adaptive inference strategies that dynamically balance computational cost or latency with predictive accuracy are critical in edge computing scenarios. In this work, we build on Adaptive Deep Neural Networks (ADNNs) that employ the Multi-Armed Bandit (MAB) framework. Current literature leverages the first version of the Upper Confidence Bound (UCB1) strategy to dynamically select the optimal confidence threshold, enabling efficient early exits without sacrificing accuracy. However, we introduce four additional Upper Confidence Bound strategies in ADNNs, namely UCB-V, UCB-Tuned, UCB-Bayes, and UCB-BwK, and perform, for the first time, a comparative study of these strategies with respect to trade-offs between accuracy, energy consumption, and latency. The proposed UCB strategies are employed on the ResNet and MobileViT neural networks, and are evaluated on the benchmark datasets of CIFAR-10, CIFAR-10.1, and CIFAR-100. Experimental results demonstrate that all strategies achieve sub-linear cumulative regret, with UCB-Bayes converging the fastest, followed by UCB-Tuned and UCB-V. Finally, UCB-V and UCB-Tuned dominate the Pareto Frontiers of accuracy-latency and accuracy-energy trade-offs.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
BARRED: Synthetic Training of Custom Policy Guardrails via Asymmetric Debate
arXiv:2604.25203v1 Announce Type: new
Abstract: Deploying guardrails for custom policies remains challenging, as generic safety models fail to capture task-specific requirements, while prompting LLMs suffers from inconsistent boundary-case performance and high inference costs. Training custom classifiers achieves both accuracy and efficiency, yet demands substantial labeled data that is costly to obtain. We present BARRED (Boundary Alignment Refinement through REflection and Debate), a framework for generating faithful and diverse synthetic training data using only a task description and a small set of unlabeled examples. Our approach decomposes the domain space into dimensions to ensure comprehensive coverage, and employs multi-agent debate to verify label correctness, yielding a high-fidelity training corpus. Experiments across diverse custom policies demonstrate that small language models finetuned on our synthetic data consistently outperform state-of-the-art proprietary LLMs (including reasoning models) and dedicated guardrail models. Ablation studies confirm that both dimension decomposition and debate-based verification are critical for ensuring the diversity and label fidelity required for effective fine-tuning. The BARRED framework eliminates the reliance on extensive human annotation, offering a scalable solution for accurate custom guardrails.
Fonte: arXiv cs.CL
NLP/LLMs • Score 92
WhisperPipe: A Resource-Efficient Streaming Architecture for Real-Time Automatic Speech Recognition
arXiv:2604.25611v1 Announce Type: new
Abstract: Real-time automatic speech recognition (ASR) systems face a fundamental trade-off between transcription accuracy and computational efficiency, particularly when deploying large-scale transformer models like Whisper. Existing streaming approaches either sacrifice accuracy through aggressive chunking or incur prohibitive memory costs through unbounded context accumulation. We present WhisperPipe, a novel streaming architecture that achieves bounded memory consumption while maintaining transcription quality through three key innovations a hybrid Voice Activity Detection (VAD) pipeline combining Silero VAD with energy-based filtering to reduce false activations by 34%, a dynamic buffering mechanism with overlapping context windows that prevents information loss at segment boundaries, and an adaptive processing strategy that balances latency and accuracy based on speech characteristics. Evaluated on 2.5 hours of diverse audio data, WhisperPipe demonstrates a median end-to-end latency of 89ms (90th percentile: 142ms) while consuming 48% less peak GPU memory and 80.9% lower average GPU utilization compared to baseline Whisper implementations. The system maintains stable memory usage over extended sessions, with zero growth rate across 150-minute continuous operation. Comparative analysis against related work shows that WhisperPipe achieves competitive accuracy (WER within 2% of offline Whisper) while operating at 3-5x lower latency than existing streaming solutions. The architecture's modular design enables deployment across resource-constrained environments, from edge devices to cloud infrastructure. Our results demonstrate that careful architectural design can reconcile the competing demands of real-time responsiveness and model sophistication in production ASR systems.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
Don\'t Stop Early: Scalable Enterprise Deep Research with Controlled Information Flow and Evidence-Aware Termination
arXiv:2604.24978v1 Announce Type: new
Abstract: Enterprise deep research often fails to produce decision-ready reports due to uneven information coverage, context explosion, and premature stopping. We propose a scalable Enterprise Deep Research (EDR) architecture to address these failures. Our system (i) decomposes requests into coverage-driven objectives via outline generation with reflection, (ii) localizes context with dependency-guided execution and explicit information sharing, and (iii) enforces evidence-based completion criteria so agents iteratively collect information until sufficiency conditions are met. We evaluate on an internal sales enablement task and the public DeepResearch Bench benchmark, where our proposed system design achieves the strongest overall performance compared with competitive deep-research baselines. The results show that dependency-controlled context and explicit evidence sufficiency criteria reduce premature stopping and improve the consistency and depth of enterprise research outputs.
Fonte: arXiv cs.CL
Applications • Score 85
Detecção automatizada de doenças cardíacas congênitas pediátricas a partir de fonocardiogramas usando fusão de características profundas e artesanais
arXiv:2604.24767v1 Tipo de Anúncio: novo
Resumo: A doença cardíaca congênita (CHD) é o tipo mais comum de defeito de nascimento, afetando cerca de 1% dos nascimentos vivos em todo o mundo. A ecocardiografia, o método diagnóstico padrão-ouro, é cara e inacessível em ambientes com poucos recursos. O diagnóstico é atrasado devido à escassez de especialistas qualificados, cuja capacidade de interpretar padrões patológicos varia significativamente, causando variabilidade inter e intra-clínica. Portanto, apresentamos um novo método para uma modalidade de diagnóstico mais acessível, o estetoscópio digital, para detectar CHDs. Nosso método é baseado na fusão de características profundas, integrando características profundas e artesanais para a detecção automatizada precoce de CHDs. Para este trabalho, gravações de Fonocardiografia (PCG) foram obtidas de 751 sujeitos pediátricos (Idade: 1 mês - 16 anos) em Bangladesh, variando de bebês a adultos em quatro locais de ausculta: válvula mitral (MV), válvula aórtica (AV), válvula pulmonar (PV) e válvula tricúspide (TV). Essas gravações foram rotuladas com base em diagnósticos confirmados por cardiologistas como casos de CHD ou não-CHD. Os resultados demonstraram que nosso modelo proposto alcançou uma precisão de 92%, uma sensibilidade de 91% e uma especificidade de 91%, com base em uma divisão de pacientes de 70% para treinamento, 20% para validação e 10% para teste. Além disso, a Área Sob a Curva da Característica Operacional do Receptor (AUROC) foi de 96%, e um F1-score de 92%. Este modelo promete uma detecção remota eficiente e em tempo real de CHDs como uma ferramenta de triagem econômica para ambientes com poucos recursos.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
PolyKV: A Shared Asymmetrically-Compressed KV Cache Pool for Multi-Agent LLM Inference
arXiv:2604.24971v1 Announce Type: new
Abstract: We present PolyKV, a system in which multiple concurrent inference agents share a single, asymmetrically compressed KV cache pool. Rather than allocating a separate KV cache per agent -- the standard paradigm -- PolyKV writes a compressed cache once and injects it into N independent agent contexts via HuggingFace DynamicCache objects. Compression is asymmetric: Keys are quantized at int8 (q8_0) to preserve softmax stability, while Values are compressed using TurboQuant MSE -- a Fast Walsh-Hadamard Transform (FWHT) rotation followed by 3-bit Lloyd-Max quantization with centroids tuned to N(0,1). We evaluate across two model scales (SmolLM2-1.7B-Instruct and Llama-3-8B-Instruct), three context lengths (600-7,194 tokens), and up to 15 concurrent agents. PolyKV achieves a stable 2.91x compression ratio across all configurations. On Llama-3-8B with 15 agents sharing a 4K-token context, PolyKV reduces KV cache memory from 19.8 GB to 0.45 GB -- a 97.7% reduction -- while maintaining only +0.57% perplexity degradation and a mean BERTScore F1 of 0.928. PPL delta does not grow with agent count and improves as context length increases, inverting to -0.26% at 1,851 coherent tokens. To our knowledge, no prior work combines a single shared, lossy-compressed KV pool with multi-reader concurrent agent access.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Why Search When You Can Transfer? Amortized Agentic Workflow Design from Structural Priors
arXiv:2604.25012v1 Announce Type: new
Abstract: Automated agentic workflow design currently relies on per-task iterative search, which is computationally prohibitive and fails to reuse structural knowledge across tasks. We observe that optimized workflows converge to a small family of domain-specific topologies, suggesting that this combinatorial search is largely redundant. Building on this insight, we propose SWIFT (Synthesizing Workflows via Few-shot Transfer), a framework that amortizes workflow design into reusable structural priors. SWIFT first distills compositional heuristics and output-interface contracts from contrastive analysis of prior search trajectories across source tasks. At inference time, it conditions a single LLM generation pass on these priors together with cross-task workflow demonstrations to synthesize a complete, executable workflow for an unseen target task, bypassing iterative search entirely. On five benchmarks, SWIFT outperforms the state-of-the-art search-based method while reducing marginal per-task optimization cost by three orders of magnitude. It further generalizes to four additional unseen benchmarks and transfers successfully from GPT-4o-mini to three additional foundation models (Grok, Qwen, Gemma). Controlled ablations reveal that workflow demonstrations primarily transfer topological structure rather than surface semantics: replacing all operator names with random strings still retains over 93% of the full system's average performance.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
minAction.net: Energy-First Neural Architecture Design -- From Biological Principles to Systematic Validation
arXiv:2604.24805v1 Announce Type: new
Abstract: Modern machine learning optimizes for accuracy without explicitly accounting for internal computational cost, even though physical and biological systems operate under intrinsic energy constraints. We evaluate energy-aware learning across 2,203 experiments spanning vision, text, neuromorphic, and physiological datasets, using 10 seeds per configuration and performing a factorial statistical analysis. Three findings emerge. First, architecture alone explains negligible variance in accuracy (partial eta^2 = 0.001). In contrast, the architecture x dataset interaction is large (partial eta^2 = 0.44, p < 0.001), demonstrating that optimal architecture depends critically on task modality and rejecting the assumption of a universal best architecture. Second, a controlled lambda-sweep over four orders of magnitude validates a single-parameter energy-regularized objective L = L_CE + lambda * E(theta, x): internal activation energy decreases to 6% of baseline at moderate lambda with no accuracy degradation on MNIST. Third, energy-first architectures inspired by an action-principle framework yield 5-33% within-modality training-efficiency gains over conventional baselines. These results emerge from a research program that interprets learning through a structural correspondence between the action functional in classical mechanics, free energy in statistical physics, and KL-regularized objectives in variational inference. We frame this correspondence as a design hypothesis rather than a derivation.
Fonte: arXiv cs.LG
Multimodal • Score 85
Nemotron 3 Nano Omni: Efficient and Open Multimodal Intelligence
arXiv:2604.24954v1 Announce Type: new
Abstract: We introduce Nemotron 3 Nano Omni, the latest model in the Nemotron multimodal series and the first to natively support audio inputs alongside text, images, and video. Nemotron 3 Nano Omni delivers consistent accuracy improvements over its predecessor, Nemotron Nano V2 VL, across all modalities, enabled by advances in architecture, training data and recipes. In particular, Nemotron 3 delivers leading results in real-world document understanding, long audio-video comprehension, and agentic computer use. Built on the highly efficient Nemotron 3 Nano 30B-A3B backbone, Nemotron 3 Nano Omni further incorporates innovative multimodal token-reduction techniques to deliver substantially lower inference latency and higher throughput than other models of similar size. We are releasing model checkpoints in BF16, FP8, and FP4 formats, along with portions of the training data and codebase to facilitate further research and development.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
CroSearch-R1: Better Leveraging Cross-lingual Knowledge for Retrieval-Augmented Generation
arXiv:2604.25182v1 Announce Type: new
Abstract: A multilingual collection may contain useful knowledge in other languages to supplement and correct the facts in the original language for Retrieval-Augmented Generation (RAG). However, the vanilla approach that simply concatenates multiple pieces of knowledge from different languages into the context may fail to improve effectiveness due to the potential disparities across languages. To better leverage multilingual knowledge, we propose CroSearch-R1, a search-augmented reinforcement learning framework to integrate multilingual knowledge into the Group Relative Policy Optimization (GRPO) process. In particular, the approach adopts a multi-turn retrieval strategy with cross-lingual knowledge integration to dynamically align the knowledge from other languages as supplementary evidence into a unified representation space. Furthermore, we introduce a multilingual rollout mechanism to optimize reasoning transferability across languages. Experimental results demonstrate that our framework effectively leverages cross-lingual complementarity and improves the effectiveness of RAG with multilingual collections.
Fonte: arXiv cs.CL
Theory/Optimization • Score 85
CoreFlow: Low-Rank Matrix Generative Models
arXiv:2604.24959v1 Announce Type: new
Abstract: Learning matrix-valued distributions from high-dimensional and possibly incomplete training data is challenging: ambient-space generative modeling is computationally expensive and statistically fragile when the matrix dimension is large but the sample size is limited. We propose CoreFlow, a geometry-preserving low-rank flow model that learns shared row/column subspaces across the matrix distribution, and then trains a continuous normalizing flow only on the induced low-dimensional core. CoreFlow is designed for settings where shared low-rank matrix geometry is present, especially in high-dimensional limited-sample regimes. This separates shared matrix geometry from sample-specific variation, preserves matrix structure, and substantially improves training efficiency. The same framework also handles incomplete training matrices through masked Riemannian updates and iterative completion. Across real and synthetic benchmarks, CoreFlow substantially improves spectral and moment-level generation quality in few-sample regimes while remaining competitive in data-rich settings, even under compression to 9% of the ambient dimension and with up to 40% missing training entries.
Fonte: arXiv cs.LG
Vision • Score 85
Contrastive Image-Metadata Pre-Training for Materials Transmission Electron Microscopy
arXiv:2604.24909v1 Announce Type: new
Abstract: The vast majority of transmission electron microscopy (TEM) data never gets published and ends up on a backup drive until deleted to free up space. These left-over datasets are rich in detail and variation, often paired with automatically saved metadata of instrument state and acquisition parameters. In this work, we introduce a dataset of 7,330 high-angle annular dark-field scanning-TEM (HAADF-STEM) images from a single instrument to learn a joint embedding space between image metadata and HAADF image. These embeddings link image style with acquisition parameters, which allows us to train a generative style transfer network that can convert experimental images into the style they would have had if they were recorded with different instrument parameters. We evaluate the performance of the network and explore the usefulness of the technique for physical denoising.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Intrinsic Mutual Information as a Modulator for Preference Optimization
arXiv:2604.24804v1 Announce Type: new
Abstract: Offline preference optimization methods, such as Direct Preference Optimization (DPO), offer significant advantages in aligning Large Language Models (LLMs) with human values. However, achieving optimal performance with these methods typically involves additional hyperparameter tuning, resulting in substantial time overhead. Although prior work has proposed a range of improvements, these methods remain limited in effectiveness and have not fully eliminated reliance on hyperparameter tuning. In this work, we propose RMiPO, a lightweight and efficient framework for offline preference optimization. RMiPO leverages intrinsic Response-level Mutual information for Preference Optimization with hyperparameter modulation, dynamically decoupling preference contributions at negligible additional computational cost. Extensive experimental results demonstrate that RMiPO achieves consistently superior performance over existing methods while reducing training overhead by more than 15\%. Our code is available at https://github.com/liavonpenn/rmipo.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
A Systematic Approach for Large Language Models Debugging
arXiv:2604.23027v1 Announce Type: new
Abstract: Large language models (LLMs) have become central to modern AI workflows, powering applications from open-ended text generation to complex agent-based reasoning. However, debugging these models remains a persistent challenge due to their opaque and probabilistic nature and the difficulty of diagnosing errors across diverse tasks and settings. This paper introduces a systematic approach for LLM debugging that treats models as observable systems, providing structured, model-agnostic methods from issue detection to model refinement. By unifying evaluation, interpretability, and error-analysis practices, our approach enables practitioners to iteratively diagnose model weaknesses, refine prompts and model parameters, and adapt data for fine-tuning or assessment, while remaining effective in contexts where standardized benchmarks and evaluation criteria are lacking. We argue that such a structured methodology not only accelerates troubleshooting but also fosters reproducibility, transparency, and scalability in the deployment of LLM-based systems.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
AutoCompress: Critical Layer Isolation for Efficient Transformer Compression
arXiv:2604.22786v1 Announce Type: new
Abstract: We present AutoCompress, a transformer compression method motivated by an empirical finding: in small transformers, Layer 0 carries disproportionately high task-critical information, with an NTK-based importance score of 3.6 compared to a maximum of 0.054 for all other layers -- a gap of over 60x. Based on this finding, we propose Critical Layer Isolation (CLI), an architecture that protects Layer 0 at full dimensionality, compresses all intermediate layers through a learned bottleneck, and restores the full dimension at the final layer. Applied to GPT-2 Medium (354.8M parameters), CLI-GPT2 achieves 204.5 perplexity on WikiText-103 with only 143.8M parameters -- a 2.47x compression ratio and 59.5% parameter reduction. Crucially, an ablation study demonstrates that a uniform bottleneck baseline of comparable size achieves only 571.8 perplexity under identical training conditions, confirming that the architectural decision to protect Layer 0 -- rather than simply reducing model size -- is the primary driver of performance. Code and checkpoints are publicly available.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Escher-Loop: Mutual Evolution by Closed-Loop Self-Referential Optimization
arXiv:2604.23472v1 Announce Type: new
Abstract: While recent autonomous agents demonstrate impressive capabilities, they predominantly rely on manually scripted workflows and handcrafted heuristics, inherently limiting their potential for open-ended improvement. To address this, we propose Escher-Loop, a fully closed-loop framework that operationalizes the mutual evolution of two distinct populations: Task Agents that solve concrete problems, and Optimizer Agents that recursively refine both the task agents and themselves. To sustain this self-referential evolution, we propose a dynamic benchmarking mechanism that seamlessly reuses the empirical scores of newly generated task agents as relative win-loss signals to update optimizers' scores. This mechanism leverages the evolution of task agents as an inherent signal to drive the evaluation and refinement of optimizers without additional overhead. Empirical evaluations on mathematical optimization problems demonstrate that Escher-Loop effectively pushes past the performance ceilings of static baselines, achieving the highest absolute peak performance across all evaluated tasks under matched compute. Remarkably, we observe that the optimizer agents dynamically adapt their strategies to match the shifting demands of high-performing task agents, which explains the system's continuous improvement and superior late-stage performance.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Avionic Main Fuel Pump Simulation and Fault-Diagnosis Benchmark
arXiv:2604.22869v1 Announce Type: new
Abstract: In many cyber-physical systems, especially in critical applications such as aeroplanes, data to train anomaly detection and diagnosis algorithms is lacking due to data protection issues and partial observability. To combat this inherent lack of data, we introduce a high-fidelity, physics-informed co-simulation of a common aircraft main-fuel-pump system modelled in \textsc{MATLAB/Simulink Simscape Fluids}. We also describe its generated time-series data with health and fault mode annotations. To show feasibility of our benchmark, we apply an unsupervised Recurrent Variational Autoencoder (RNN-VAE) for anomaly detection and a SOM-VAE for operating mode discretization, trained to separate healthy and faulty conditions.
Fonte: arXiv cs.LG
RL • Score 85
CODA: Coordination via On-Policy Diffusion for Multi-Agent Offline Reinforcement Learning
arXiv:2604.23308v1 Announce Type: new
Abstract: Offline multi-agent reinforcement learning (MARL) enables policy learning from fixed datasets, but is prone to coordination failure: agents trained on static, off-policy data converge to suboptimal joint behaviours because they cannot co-adapt as their policies change. We introduce CODA (Coordination via On-Policy Diffusion for Multi-Agent Reinforcement Learning), a diffusion-based multi-agent trajectory generator for data augmentation that samples conditioned on the current joint policy, producing synthetic experience which reflects the evolving behaviours of the agents, thereby providing a mechanism for co-adaptation. We find that previous diffusion-based augmentation approaches are insufficient for fostering multi-agent coordination because they produce static augmented datasets that do not evolve as the current joint policy changes during training; CODA resolves this by more closely simulating on-policy learning and is a meaningful step toward coordinated behaviours in the offline setting. CODA is algorithm-agnostic and can be layered onto both model-free and model-based offline reinforcement learning pipelines as an augmentation module. Empirically, CODA not only resolves canonical coordination pathologies in continuous polynomial games but also delivers strong results on the more complex MaMuJoCo continuous-control benchmarks.
Fonte: arXiv cs.LG
MLOps/Systems • Score 75
Efficient VQ-QAT and Mixed Vector/Linear quantized Neural Networks
arXiv:2604.23172v1 Announce Type: new
Abstract: In this work, we developed and tested 3 techniques for vector quantization (VQ) based model weight compression. To mitigate codebook collapse and enable end-to-end training, we adopted cosine similarity-based assignment. Building on ideas from attention-based formulations in Differentiable K-Means (DKM), we further improved this approach by using cosine similarity for assignment combined with top-1 sampling and a straight-through estimator, thereby eliminating the need for weighted-average reconstruction. Finally, we investigated the use of differentiable neural architecture search (NAS) to adaptively select layer-wise quantization configurations, further optimizing the compression process. Although our method does not consistently outperform existing approaches across all quantization levels, it provides useful insights into the design trade-offs and behaviors of VQ-based model compression methods.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
ClawTrace: Cost-Aware Tracing for LLM Agent Skill Distillation
arXiv:2604.23853v1 Announce Type: new
Abstract: Skill-distillation pipelines learn reusable rules from LLM agent trajectories, but they lack a key signal: how much each step costs. Without per-step cost, a pipeline cannot distinguish adding a missing step to fix a bug from removing an expensive step that never affected the outcome. We introduce ClawTrace, an agent tracing platform that records every LLM call, tool use, and sub-agent spawn during an agent session and compiles each session into a TraceCard: a compact YAML summary with per-step USD cost, token counts, and redundancy flags. Built on ClawTrace, CostCraft is a distillation pipeline that reads TraceCards and produces three types of skill patches. Preserve patches keep behaviors that led to success. Prune patches remove expensive steps that did not matter, each backed by a counterfactual argument against a named high-cost step. Repair patches fix failures grounded in oracle evidence. Ablations on 30 held-out SpreadsheetBench tasks show that both cost attribution and prune patches independently reduce quality regressions. When the same skill is applied to 30 unrelated SkillsBench tasks, an unexpected asymmetry emerges: prune rules transferred across benchmarks and cut median cost by 32%, while preserve rules, trained on benchmark-specific conventions, caused regressions on new task types. We release ClawTrace and TraceCards as open infrastructure for cost-aware agent research.
Fonte: arXiv cs.AI
NLP/LLMs • Score 90
Mixture of Heterogeneous Grouped Experts for Language Modeling
arXiv:2604.23108v1 Announce Type: new
Abstract: Large Language Models (LLMs) based on Mixture-of-Experts (MoE) are pivotal in industrial applications for their ability to scale performance efficiently. However, standard MoEs enforce uniform expert sizes,creating a rigidity that fails to align computational costs with varying token-level complexity. While heterogeneous expert architectures attempt to address this by diversifying expert sizes, they often suffer from significant system-level challenges, specifically unbalanced GPU utilization and inefficient parameter utilization, which hinder practical deployment. To bridge the gap between theoretical heterogeneity and robust industrial application, we propose Mixture of Heterogeneous Grouped Experts (MoHGE) which introduces a two-level routing mechanism to enable flexible, resource-aware expert combinations. To optimize inference efficiency, we propose a Group-Wise Auxiliary Loss, which dynamically steers tokens to the most parameter-efficient expert groups based on task difficulty. To address the critical deployment challenge of GPU load balancing, we introduce an All-size Group-decoupling Allocation strategy coupled with an Intra-Group Experts Auxiliary Loss. These mechanisms collectively ensure uniform computation distribution across GPUs. Extensive evaluations demonstrate that MoHGE matches the performance of MoE architectures while reducing the total parameters by approximately 20% and maintaining balanced GPU utilization. Our work establishes a scalable paradigm for resource-efficient MoE design, offering a practical solution for optimizing inference costs in real-world scenarios.
Fonte: arXiv cs.CL
RL • Score 85
CoFi-PGMA: Counterfactual Policy Gradients under Filtered Feedback for Multi-Agent LLMs
arXiv:2604.22785v1 Announce Type: new
Abstract: Large language model (LLM) deployments increasingly rely on multi-agent architectures in which multiple models either compete through routing mechanisms or collaborate to produce a final answer. In both settings, the learning signal received by each agent is filtered by the system mechanism. Routing produces selection-gated feedback where only the chosen response is evaluated, while collaboration produces shared rewards that obscure the individual contribution of each agent. As a result, standard RLHF objectives designed for a single deployed policy become misspecified. We introduce CoFi-PGMA (Counterfactual Policy Gradients under Filtered Feedback for Multi-Agent LLMs), a unified framework for learning under filtered feedback in multi-agent LLM systems. Our approach derives a counterfactual per-agent training objective based on marginal contribution, which corrects the learning signal under both routing and collaborative mechanisms. For routing systems, the objective corresponds to off-policy corrections for selection-gated feedback, while for collaborative systems it reduces to leave-one-out difference rewards for credit assignment. We further analyze how softmax routing induces risk-sensitive incentives and provide practical training algorithms that integrate counterfactual estimators, multiturn-aware rewards, and policy optimization methods, and demonstrate the approach on a real-world reasoning dataset.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
An Intelligent Fault Diagnosis Method for General Aviation Aircraft Based on Multi-Fidelity Digital Twin and FMEA Knowledge Enhancement
arXiv:2604.22777v1 Announce Type: new
Abstract: Fault diagnosis of general aviation aircraft faces challenges including scarce real fault data, diverse fault types, and weak fault signatures. This paper proposes an intelligent fault diagnosis framework based on multi-fidelity digital twin, integrating four modules: high-fidelity flight dynamics simulation, FMEA-driven fault injection, multi-fidelity residual feature extraction, and large language model (LLM)-enhanced interpretable report generation. A digital twin is constructed using the JSBSim six-degree-of-freedom (6-DoF) flight dynamics engine, generating 23-channel engine health monitoring data via semi-empirical sensor synthesis equations. A three-layer fault injection engine based on failure mode and effects analysis (FMEA) models the physical causal propagation of 19 engine fault types. A multi-fidelity residual computation framework comprising paired-mirror residuals and GRU surrogate prediction residuals is proposed: the high-fidelity path obtains clean fault deviation signals using nominal mirror trajectories with identical initial conditions, while the low-fidelity path achieves online real-time residual computation through a multi-step prediction GRU surrogate model. A 1D-CNN classifier performs end-to-end diagnosis of 20 fault classes. An LLM diagnostic report engine enhanced with FMEA knowledge fuses classification results, residual evidence, and domain causal knowledge to generate interpretable natural language reports. Experiments show the paired-mirror residual scheme achieves a Macro-F1 of 96.2% on the 20-class task, while the GRU surrogate scheme achieves 4.3x inference acceleration at only 0.6% performance cost. Comparison across 24 schemes reveals that residual feature quality contributes approximately 5x more to diagnostic performance than classifier architecture, establishing the "residual quality first" design principle.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
Shape of Memory: a Geometric Analysis of Machine Unlearning in Second-Order Optimizers
arXiv:2604.23046v1 Announce Type: new
Abstract: We argue that current definitions of machine unlearning are underspecified for second-order optimizers. We compare first-order and second-order learners for their ability to handle the data deletion task with varying degrees of eigendecomposition to mimic the loss model memory. While both first and second-order methods realign with the ideal counterfactul in terms of performance and gradient, the second-order optimizer shows significant volatility in the optimizer state. This indicates residual information, supposedly deleted, that isn't detectable by first-order analysis. Various eigendecay treatments show that stability and information loss is regained only under controlled state pertubation where geometric information (or memory) is erased.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Stochastic KV Routing: Enabling Adaptive Depth-Wise Cache Sharing
arXiv:2604.22782v1 Announce Type: new
Abstract: Serving transformer language models with high throughput requires caching Key-Values (KVs) to avoid redundant computation during autoregressive generation. The memory footprint of KV caching is significant and heavily impacts serving costs. This work proposes to lessen these memory requirements. While recent work has largely addressed KV cache reduction via compression and eviction along the temporal axis, we argue that the \emph{depth} dimension offers an orthogonal and robust avenue for optimization. Although prior research suggests that a full cache for every layer is redundant, implementing cross-layer cache sharing remains a practical challenge; existing methods typically suffer from reduced throughput or increased time-to-first-token. In this paper, we demonstrate that dropping a layer's cache offers efficient optimization without information loss. We propose a simple training approach: random cross-layer attention. During training, layers randomly choose to attend either to their own KV states or those of a preceding layer. This stochastic process adapts the model to be robust to various depth-wise cache sharing strategies, ensuring flexibility for unknown hardware constraints at deployment time. Our evaluations show that applying this scheme during pre-training or fine-tuning enables depth-wise cache sharing for various model families. Furthermore, for larger models in data-constrained settings, this approach is suggestive of a regularization-like effect, frequently preserving or improving performance while significantly reducing the cache's memory footprint.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Vibe Medicine: Redefining Biomedical Research Through Human-AI Co-Work
arXiv:2604.23674v1 Announce Type: new
Abstract: With the emergence of large language models (LLMs) and AI agent frameworks, the human-AI co-work paradigm known as Vibe Coding is changing how people code, making it more accessible and productive. In scientific research, where workflows are more complex and the burden of specialized labor limits independent researchers and those in low-resource areas, the potential impact is even greater, particularly in biomedicine, which involves heterogeneous data modalities and multi-step analytical pipelines. In this paper, we introduce Vibe Medicine, a co-work paradigm in which clinicians and researchers direct skill-augmented AI agents through natural language to execute complex, multi-step biomedical workflows, while retaining the role of research director who specifies objectives, reviews intermediate results, and makes domain-informed decisions. The enabling infrastructure consists of three layers: capable LLMs, agent frameworks such as OpenClaw and Hermes Agent, and the OpenClaw medical skills collection, which includes more than 1,000 curated skills from multiple open-source repositories. We analyze the architecture and skill categories of this collection across ten biomedical domains, and present case studies covering rare disease diagnosis, drug repurposing, and clinical trial design that demonstrate end-to-end workflows in practice. We also identify the principal risks, such as hallucination, data privacy, and over-reliance, and outline directions toward more reliable, trustworthy, and clinically integrated agent-assisted research that advances research and technological equity and reduces health care resource disparities.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Time-Series Forecasting in Safety-Critical Environments: An EU-AI-Act-Compliant Open-Source Package / Zeitreihenprognose in sicherheitskritischen Umgebungen: Ein KI-VO-konformes Open-Source-Paket
arXiv:2604.23859v1 Announce Type: new
Abstract: With spotforecast2-safe we present an integrated Compliance-by-Design approach to Python-based point forecasting of time series in safety-critical environments. A review of the relevant open-source tooling shows that existing compliance solutions operate consistently outside of the library to be used - e.g. as scanners, templates, or runtime layers. spotforecast2-safe takes the inverse approach and anchors the requirements of Regulation (EU) 2024/1689 (the EU AI Act, in German: KI-VO), of IEC 61508, of the ISA/IEC 62443 standards series, and of the Cyber Resilience Act within the library: in application-programming-interface contracts, persistence formats, and continuous-integration gates. The approach is operationalised by four non-negotiable code-development rules (zero dead code, deterministic processing, fail-safe handling, minimal dependencies) together with the corresponding process rules (model card, executable docstrings, CI workflows, Common-Platform-Enumeration (CPE) identifier, REUSE-conformant licensing, release pipeline). Interactive visualisation, hyperparameter tuning and automated machine learning (AutoML), as well as deep-learning and large-language-model backends are deliberately excluded, because each of these components either enlarges the attack surface, introduces non-determinism, or impairs reproducibility. A bidirectional traceability matrix maps every regulatory provision onto the corresponding mechanism in the code; an end-to-end example of European-market electricity generation, transmission, and consumption forecasting demonstrates the application. The package is open-source and available under Affero General Public License (AGPL) 3.0-or-later.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
A Decoupled Human-in-the-Loop System for Controlled Autonomy in Agentic Workflows
arXiv:2604.23049v1 Announce Type: new
Abstract: AI agents are increasingly deployed to execute tasks and make decisions within agentic workflows, introducing new requirements for safe and controlled autonomy. Prior work has established the importance of human oversight for ensuring transparency, accountability, and trustworthiness in such systems. However, existing implementations of Human-in-the-Loop (HITL) mechanisms are typically embedded within application logic, limiting reuse, consistency, and scalability across multi-agent environments.
This paper presents a decoupled HITL system architecture that treats human oversight as an independent system component within the agent operating environment. The proposed design separates human interaction management from application workflows through explicit interfaces and a structured execution model. In addition, a design framework is introduced to formalize HITL integration along four dimensions: intervention conditions, role resolution, interaction semantics, and communication channel. This framework enables selective and context-aware human involvement while maintaining system-level consistency.
The approach supports alignment with emerging agent communication protocols, allowing HITL to be implemented as a protocol-level concern. By externalizing HITL and structuring its integration, the system provides a foundation for scalable governance and progressive autonomy in agentic workflows.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
IndustryAssetEQA: A Neurosymbolic Operational Intelligence System for Embodied Question Answering in Industrial Asset Maintenance
arXiv:2604.23446v1 Announce Type: new
Abstract: Industrial maintenance environments increasingly rely on AI systems to assist operators in understanding asset behavior, diagnosing failures, and evaluating interventions. Although large language models (LLMs) enable fluent natural-language interaction, deployed maintenance assistants routinely produce generic explanations that are weakly grounded in telemetry, omit verifiable provenance, and offer no testable support for counterfactual or action-oriented reasoning that undermine trust in safety-critical settings. We present IndustryAssetEQA, a neurosymbolic operational intelligence system that combines episodic telemetry representations with a Failure Mode Effects Analysis Knowledge Graph (FMEA-KG) to enable Embodied Question Answering (EQA) over industrial assets. We evaluate on four datasets covering four industrial asset types, including rotating machinery, turbofan engines, hydraulic systems, and cyber-physical production systems. Compared to LLM-only baselines, IndustryAssetEQA improves structural validity by up to 0.51, counterfactual accuracy by up to 0.47, and explanation entailment by 0.64, while reducing severe expert-rated overclaims from 28% to 2% (approximately 93% reduction). Code, datasets, and the FMEA-KG are available at https://github.com/IBM/AssetOpsBench/tree/IndustryAssetEQA/IndustryAssetEQA.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
LEGO: An LLM Skill-Based Front-End Design Generation Platform
arXiv:2604.23355v1 Announce Type: new
Abstract: Existing LLM-based EDA agents are often isolated task-specific systems. This leads to repeated engineering effort and limited reuse of successful design and debugging strategies. We present LEGO, a unified skill-based platform for front-end design generation. It decomposes the digital front-end flow into six independent steps and represents every agent capability as a standardized composable circuit skill within a plug-and-play architecture. To build this skill library, we survey more than 100 papers, select 11 representative open-source projects, and extract 42 executable circuit skills within a six-step finite state machine formulation. Circuit Skill Builder automates skill extraction with linear scalability. Agent Skill RAG achieves submillisecond retrieval without relying on embedding models. Empirical evaluation on a hard subset of 41 VerilogEval v2 problems that gpt-5.2-codex fails to solve under extra-high reasoning effort shows that individual circuit skills constructed within LEGO raise Pass@1 from 0.000 to 0.805. This is an 80.5% gain over the baseline. Cross-project skill compositions also reach 0.805 Pass@1. They outperform hierarchy-verilog by 14.6% and VerilogCoder by 2.5%. They also match MAGE. These results show that modular skill composition supports both effective and flexible RTL design automation. The LEGO platform and all circuit skills are publicly available at GitHub: https://github.com/loujc/LEGO-An-LLM-Skill-Based-Front-End-Design-Generation-Platform
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Learning Without Adversarial Training: A Physics-Informed Neural Network for Secure Power System State Estimation under False Data Injection Attacks
arXiv:2604.22784v1 Announce Type: new
Abstract: State estimation is a cornerstone of power system control-center operations, and its robust operation is increasingly a cyber-physical security concern as modern grids become more digitalized and communication-intensive. Neural network-based approaches have gained attention as alternatives to conventional model-based state estimation methods. Physics-Informed Neural Networks (PINNs), which embed power-flow consistency into the learning objective, have shown improved accuracy over existing approaches. This work proposes a PINN-based model for Power System State Estimation (PSSE) that protects the estimation process against the stealth-constrained AC False Data Injection Attacks (FDIAs) considered in this study. The model is developed without adversarial training. Instead, a dynamic loss-weighting formulation based on homoscedastic uncertainty learns the relative scaling of supervised data-fit and physics-residual terms during training, reducing sensitivity to manual weight tuning. Robustness is evaluated on the IEEE 118-bus system using representative stealthy-FDIA families including state distortion, load redistribution, line overloading, and residual-constrained stealth corruption. Performance is measured using Mean Absolute Error (MAE) on voltage magnitudes and phase angles. Results demonstrate higher accuracy and stability than existing fixed-weight PINN variants.
Fonte: arXiv cs.LG
NLP/LLMs • Score 92
GraphPlanner: Graph Memory-Augmented Agentic Routing for Multi-Agent LLMs
arXiv:2604.23626v1 Announce Type: new
Abstract: LLM routing has achieved promising results in integrating the strengths of diverse models while balancing efficiency and performance. However, to support more realistic and challenging applications, routing must extend into agentic LLM settings, where task planning, multi-round cooperation among heterogeneous agents, and memory utilization are indispensable. To address this gap, we propose GraphPlanner, a heterogeneous graph memory-augmented agentic router for multi-agent LLMs that generates routing workflows for each query and supports both inductive and transductive inference. GraphPlanner formulates workflow generation as a Markov Decision Process (MDP), where at each step it selects both the LLM backbone and the agent role, including Planner, Executor, and Summarizer. By leveraging a heterogeneous graph, denoted as GARNet, to capture interaction memories among queries, agents, and responses, GraphPlanner integrates historical memory and workflow memory into richer state representations. The entire pipeline is optimized with reinforcement learning, jointly improving task-specific performance and computational efficiency. We evaluate GraphPlanner across 14 diverse LLM tasks and demonstrate that: (1) GraphPlanner outperforms strong single-round and multi-round routers, improving accuracy by up to 9.3% while reducing GPU cost from 186.26 GiB to 1.04 GiB; (2) GraphPlanner generalizes robustly to unseen tasks and LLMs, exhibiting strong zero-shot capabilities; and (3) GraphPlanner effectively leverages historical memories, supporting both inductive and transductive inference for more adaptive routing. Our code for GraphPlanner is released at https://github.com/ulab-uiuc/GraphPlanner.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
RouteNLP: Closed-Loop LLM Routing with Conformal Cascading and Distillation Co-Optimization
arXiv:2604.23577v1 Announce Type: new
Abstract: Serving diverse NLP workloads with large language models is costly: at one enterprise partner, inference costs exceeded $200K/month despite over 70% of queries being routine tasks well within the capability of smaller models. We present RouteNLP, a closed-loop framework that routes queries across a tiered model portfolio to minimize cost while satisfying per-task quality constraints. The framework integrates three components: a difficulty-aware router with shared task-conditioned representations trained on preference data and quality signals; confidence-calibrated cascading that uses conformal prediction for distribution-free threshold initialization; and a distillation-routing co-optimization loop that clusters escalation failures, applies targeted knowledge distillation to cheaper models, and automatically retrains the router, yielding over twice the cost improvement of untargeted distillation. In an 8-week pilot deployment processing ~5K queries/day at an enterprise customer-service division, RouteNLP reduced inference costs by 58% while maintaining 91% response acceptance and reducing p99 latency from 1,847 ms to 387 ms. On a six-task benchmark spanning finance, customer service, and legal domains, the framework achieves 40-85% cost reduction while retaining 96-100% quality on structured tasks and 96-98% on generation tasks, with human evaluation confirming that 74.5% of routed generation outputs match or exceed frontier-model quality.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
Parameter Efficiency Is Not Memory Efficiency: Rethinking Fine-Tuning for On-Device LLM Adaptation
arXiv:2604.22783v1 Announce Type: new
Abstract: Parameter-Efficient Fine-Tuning (PEFT) has become the standard for adapting large language models (LLMs). In this work we challenge the wide-spread assumption that parameter efficiency equates memory efficiency and on-device adaptability. We show that this is not true - while methods like LoRA and IA3 significantly reduce trainable parameters, they remain bound by intermediate tensors that scale linearly with sequence length, often triggering out-of-memory errors on-device. In this work, we introduce LARS (Low-memory Activation-Rank Subspace), a novel adaptation framework that decouples memory consumption from sequence length. While prior PEFT methods apply low-rank constraints to model parameters, LARS instead constrains the activation subspace used during training, directly targeting the dominant source of memory consumption and fundamentally flattening the memory growth rate. LARS reduces the memory footprint by an average of 33.54% on GPUs and 51.95% on CPUs in comparison to LoRA across reasoning, understanding and long-context datasets using different models while maintaining competitive accuracy and throughput. Besides GPUs, we deploy on Raspberry Pi and consumer-grade CPUs to demonstrate that LARS provides a scalable path for sophisticated LLM personalization on resource-constrained hardware and edge devices.
Fonte: arXiv cs.LG
Multimodal • Score 85
SoccerRef-Agents: Multi-Agent System for Automated Soccer Refereeing
arXiv:2604.23392v1 Announce Type: new
Abstract: Refereeing is vital in sports, where fair, accurate, and explainable decisions are fundamental. While intelligent assistant technologies are being widely adopted in soccer refereeing, current AI-assisted approaches remain preliminary. Existing research mostly focuses on isolated video perception tasks and lacks the ability to understand and reason about foul scenarios. To fill this gap, we propose SoccerRef-Agents, a holistic and explainable multi-agent decision-making framework for soccer refereeing. The main contributions are: (i) constructing the multimodal benchmark SoccerRefBench with over 1,200 referee theory questions and 600 foul video clips; (ii) building a vector-based knowledge base RefKnowledgeDB using the latest "Laws of the Game" and a classic case database for precise, knowledge-driven reasoning; (iii) designing a novel multi-agent architecture that collaborates via cross-modal RAG to bridge the semantic gap between visual content and regulatory texts. This work explores the technical capability of integrating MLLMs with refereeing expertise, and evaluations show our system significantly outperforms general-purpose MLLMs in decision accuracy and explanation quality. All databases, benchmarks, and code will be made available.
Fonte: arXiv cs.AI
RecSys • Score 85
MTServe: Efficient Serving for Generative Recommendation Models with Hierarchical Caches
arXiv:2604.22881v1 Announce Type: new
Abstract: Generative recommendation (GR) offers superior modeling capabilities but suffers from prohibitive inference costs due to the repeated encoding of long user histories. While cross-request Key-Value (KV) cache reuse presents a significant optimization opportunity, the massive scale of individual user states creates a storage explosion that far exceeds physical GPU limits. We propose MTServe, a hierarchical cache management system that virtualizes GPU memory by leveraging host RAM as a scalable backup store. To bridge the I/O gap between tiers, MTServe introduces a suite of system-level optimizations, including a hybrid storage layout, an asynchronous data transfer pipeline, and a locality-driven replacement policy. On both public and production datasets, MTServe delivers up to 3.1* speedup while maintaining near-perfect hit ratios (>98.5%).
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Accelerating Frequency Domain Diffusion Models with Error-Feedback Event-Driven Caching
arXiv:2604.22901v1 Announce Type: new
Abstract: Diffusion models achieve remarkable success in time series generation. However, slow inference limits their practical deployment. We propose E$^2$-CRF (Error-Feedback Event-Driven Cumulative Residual Feature caching) to accelerate frequency domain diffusion models. Our method exploits two structural properties: (1) spectral localization, where signal energy concentrates in low frequencies, and (2) mirror symmetry, which halves the effective frequency dimension. E$^2$-CRF uses a closed-loop error-feedback system that adaptively caches transformer KV features across diffusion steps. We trigger recomputation using event-driven residual dynamics instead of fixed schedules. Our method selectively recomputes high-energy or rapidly-changing tokens while reusing cached features for stable high-frequency components. E$^2$-CRF achieves ~2.2 speedup while maintaining sample quality. We demonstrate effectiveness on 5 datasets. Our caching strategy naturally aligns with the diffusion process's structure-to-detail progression. We include sufficient-condition error and complexity bounds under standard regularity assumptions (Appendix), alongside empirical validation. Our code is available at https://github.com/NoakLiu/FastFourierDiffusion and is also integrated in https://github.com/NoakLiu/FastCache-xDiT.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Predicting Wind Loads on Container Ships in Harbor Environments through Multi-Fidelity Modeling
arXiv:2604.22882v1 Announce Type: new
Abstract: Modern container ships face higher wind loads due to increased windage areas, making accurate predictions of wind loads essential for mooring design. Existing empirical models, largely developed for container ships with smaller windage areas and simpler geometrical configurations than those of modern large-scale vessels, often lack accuracy and do not account for the influence of nearby structures. This study proposes a multi-fidelity surrogate modelling framework for the prediction of wind-load coefficients, combining empirical correlations with simplified and detailed CFD models for ships in open-sea and harbor environments. The approach relies on recursive co-kriging to consistently fuse information across fidelity levels, enabling accurate predictions at a reduced computational cost. A sensitivity analysis is used to identify the most influential geometric parameters, and the resulting reduced parameter space is explored through sequential sampling to efficiently construct the training database.
The surrogate models are validated over a wide range of loading configurations and for two distinct harbor environments. The results demonstrate that the multi-fidelity approach significantly improves prediction accuracy compared to single-fidelity models, while substantially reducing the reliance on high-fidelity simulations. In particular, the proposed framework captures the dependence of wind loads on key geometric parameters and consistently outperforms traditional empirical correlations, providing a robust and efficient tool for engineering applications.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Scaling Multi-Node Mixture-of-Experts Inference Using Expert Activation Patterns
arXiv:2604.23150v1 Announce Type: new
Abstract: Most recent state-of-the-art (SOTA) large language models (LLMs) use Mixture-of-Experts (MoE) architectures to scale model capacity without proportional per-token compute, enabling higher-quality outputs at manageable serving costs. However, MoE inference at scale is fundamentally bottlenecked by expert load imbalance and inefficient token routing, especially in multi-node deployments where tokens are not guaranteed to be routed to local experts, resulting in significant inter-node all-to-all communication overhead.
To systematically characterize these challenges, we profile SOTA open-source MoE models, including Llama 4 Maverick, DeepSeek V3-671B, and Qwen3-230B-A22B, on various datasets and collected over 100k real expert activation traces. Upon studying the expert activation patterns, we uncover various persistent properties across all the frontier MoE models: variable expert load imbalance, domain-specific expert activation where expert popularity shifts across task families (code, math, chat, general), and a strong correlation between prefill and decode expert activations. Motivated by these findings, we propose workload-aware micro-batch grouping and an expert placement strategy to maximize token locality to the destination expert, thereby reducing inter-node communication. Across models and datasets, these optimizations help reduce all2all communication data up to 20, resulting in lower MoE decode latency and better accelerator utilization.
Fonte: arXiv cs.LG
Evaluation/Benchmarks • Score 85
ProEval: Proactive Failure Discovery and Efficient Performance Estimation for Generative AI Evaluation
arXiv:2604.23099v1 Announce Type: new
Abstract: Evaluating generative AI models is increasingly resource-intensive due to slow inference, expensive raters, and a rapidly growing landscape of models and benchmarks. We propose ProEval, a proactive evaluation framework that leverages transfer learning to efficiently estimate performance and identify failure cases. ProEval employs pre-trained Gaussian Processes (GPs) as surrogates for the performance score function, mapping model inputs to metrics such as the severity of errors or safety violations. By framing performance estimation as Bayesian quadrature (BQ) and failure discovery as superlevel set sampling, we develop uncertainty-aware decision strategies that actively select or synthesize highly informative inputs for testing. Theoretically, we prove that our pre-trained GP-based BQ estimator is unbiased and bounded. Empirically, extensive experiments on reasoning, safety alignment, and classification benchmarks demonstrate that ProEval is significantly more efficient than competitive baselines. It requires 8-65x fewer samples to achieve estimates within 1% of the ground truth, while simultaneously revealing more diverse failure cases under a stricter evaluation budget.
Fonte: arXiv cs.LG
Vision • Score 85
On-Device Vision Training, Deployment, and Inference on a Thumb-Sized Microcontroller
arXiv:2604.23012v1 Announce Type: new
Abstract: This paper presents a complete, end-to-end on-device vision machine learning pipeline, comprising data acquisition, two-layer CNN training with Adam optimization, and real-time inference, executing entirely on a microcontroller-class device costing $15-40 USD. Unlike cloud-based workflows that require external infrastructure and conceal the computational pipeline from the practitioner, this system implements every step of the core ML lifecycle in approximately 1,750 lines of readable C++ that compiles in under one minute using the Arduino IDE, with no external ML dependencies. Running on the Seeed Studio ESP32-S3 XIAO ML Kit (8 MB PSRAM), the firmware achieves three-class 64x64 image classification in approximately 9 minutes per training run, with real-time inference at 6.3 FPS. Key contributions include: correct batch-level gradient accumulation; pre-computed resize lookup tables for inference; dual-format weight export for SD-free baked-in deployment; a three-tier weight priority system (SD binary > baked-in header > He-initialization) resolved automatically at boot; a single-constant network reconfiguration interface; and PSRAM-aware memory management suited to microcontroller constraints. All source code and reference datasets are released under the MIT License at https://github.com/webmcu-ai/on-device-vision-ai
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
BiTA: Bidirectional Gated Recurrent Unit-Transformer Aggregator in a Temporal Graph Network Framework for Alert Prediction in Computer Networks
arXiv:2604.22781v1 Announce Type: new
Abstract: Proactive alert prediction in computer networks is critical for mitigating evolving cyber threats and enabling timely defensive actions. Temporal Graph Neural Networks (TGNs) provide a principled framework for modeling time-evolving interactions; however, existing TGN-based methods predominantly rely on unidirectional or single-mechanism temporal aggregation, which limits their ability to capture recursive, multi-scale temporal patterns commonly observed in real-world attack behaviors. In this paper, we propose BiTA, a Bidirectional Gated Recurrent Unit-Transformer Aggregator for temporal graph learning. Rather than introducing a deeper or higher-capacity model, BiTA redesigns the temporal aggregation function within the TGN framework by jointly encoding bidirectional sequential dependencies and long-range contextual relations over each node's temporal neighborhood. This aggregation strategy enables complementary temporal reasoning at different scales while preserving the original TGN memory and message-passing structure. We evaluate BiTA on real-world alert datasets, demonstrating significant improvements in key performance metrics such as area under the curve, average precision, mean reciprocal rank, and per-category prediction accuracy when compared to state-of-the-art temporal graph models. BiTA outperforms baseline methods under both transductive and inductive settings, highlighting its robustness and generalization capabilities in dynamic network environments. BiTA is a scalable and interpretable framework for real-time cyber threat anticipation, paving the way toward more intelligent and adaptive intrusion detection systems.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
A Layer Separation Optimization Framework for Cross-Entropy Training in Deep Learning
arXiv:2604.23225v1 Announce Type: new
Abstract: This paper investigates the deep learning optimization problem with softmax cross-entropy loss. We propose a layer separation strategy to alleviate the strong nonconvexity encountered during training deep networks. For cross-entropy models with fully connected and convolutional neural networks, we introduce auxiliary variables associated with hidden layer outputs and construct corresponding layer separation models, which decompose the original deeply nested optimization problem into a sequence of more manageable subproblems. We also conduct theoretical analyses, proving that the new layer separation loss provides an upper bound for the original cross-entropy loss. Moreover, we design alternating minimization algorithms and prove that, under appropriate conditions, these algorithms exhibit decreasing properties of the loss function. Numerical experiments validate the effectiveness of the proposed methods and indicate improved optimization behavior, especially for fully connected and convolutional neural networks.
Fonte: arXiv cs.LG
Multimodal • Score 85
Conditional Imputation for Within-Modality Missingness in Multi-Modal Federated Learning
arXiv:2604.23112v1 Announce Type: new
Abstract: Multimodal Federated Learning (MMFL) enables privacy-preserving collaborative training, but real-world clinical applications often suffer from within-modality missingness caused by sensor intermittency or irregular sampling. Existing methods implicitly represent unobserved data via architectural alignment or missing embeddings, often failing to recover the true distribution and yielding sub-optimal performance. We propose CondI, a federated framework explicitly addressing this missingness using conditional diffusion models. CondI employs a two-phase training pipeline: first, imputing unobserved temporal components using available multimodal context and conditional embeddings; second, optimizing modality-specific extractors and joint embedding spaces. During inference, imputed raw data pass through trained extractors to generate robust features, providing a holistic representation for downstream tasks. Explicit data imputation ensures models operate on complete semantic structures, significantly enhancing resilience against severe data incompleteness. Experiments on three clinical datasets (PTB-XL, SLEEP-EDF, MIMIC-IV) demonstrate CondI achieves comparable results to state-of-the-art baselines. Code: https://github.com/ZhengWugeng/CondI
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
MTRouter: Cost-Aware Multi-Turn LLM Routing with History-Model Joint Embeddings
arXiv:2604.23530v1 Announce Type: new
Abstract: Multi-turn, long-horizon tasks are increasingly common for large language models (LLMs), but solving them typically requires many sequential model invocations, accumulating substantial inference costs. Here, we study cost-aware multi-turn LLM routing: selecting which model to invoke at each turn from a model pool, given a fixed cost budget. We propose MTRouter, which encodes the interaction history and candidate models into joint history-model embeddings, and learns an outcome estimator from logged trajectories to predict turn-level model utility. Experiments show that MTRouter improves the performance-cost trade-off: on ScienceWorld, it surpasses GPT-5 while reducing total cost by 58.7%; on Humanity's Last Exam (HLE), it achieves competitive accuracy while reducing total cost by 43.4% relative to GPT-5, and these gains even carry over to held-out tasks. Further analyses reveal several mechanisms underlying its effectiveness: relative to prior multi-turn routers, MTRouter makes fewer model switches, is more tolerant to transient errors, and exhibits emergent specialization across models. Code: https://github.com/ZhangYiqun018/MTRouter
Fonte: arXiv cs.CL
Multimodal • Score 85
Focus Session: Hardware and Software Techniques for Accelerating Multimodal Foundation Models
arXiv:2604.21952v1 Announce Type: cross
Abstract: This work presents a multi-layered methodology for efficiently accelerating multimodal foundation models (MFMs). It combines hardware and software co-design of transformer blocks with an optimization pipeline that reduces computational and memory requirements. During model development, it employs performance enhancements through fine-tuning for domain-specific adaptation. Our methodology further incorporates hardware and software techniques for optimizing MFMs. Specifically, it employs MFM compression using hierarchy-aware mixed-precision quantization and structural pruning for transformer blocks and MLP channels. It also optimizes operations through speculative decoding, model cascading that routes queries through a small-to-large cascade and uses lightweight self-tests to determine when to escalate to larger models, as well as co-optimization of sequence length, visual resolution & stride, and graph-level operator fusion. To efficiently execute the model, the processing dataflow is optimized based on the underlying hardware architecture together with memory-efficient attention to meet on-chip bandwidth and latency budgets. To support this, a specialized hardware accelerator for the transformer workloads is employed, which can be developed through expert design or an LLM-aided design approach. We demonstrate the effectiveness of the proposed methodology on medical-MFMs and on code generation tasks, and conclude with extensions toward energy-efficient spiking-MFMs.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
LayerBoost: Layer-Aware Attention Reduction for Efficient LLMs
arXiv:2604.22050v1 Announce Type: new
Abstract: Transformers are mostly relying on softmax attention, which introduces quadratic complexity with respect to sequence length and remains a major bottleneck for efficient inference. Prior work on linear or hybrid attention typically replaces softmax attention uniformly across all layers, often leading to significant performance degradation or requiring extensive retraining to recover model quality.
This work proposes LayerBoost, a layer-aware attention reduction method that selectively modifies the attention mechanism based on the sensitivity of individual transformer layers. It first performs a systematic sensitivity analysis on a pretrained model to identify layers that are critical for maintaining performance. Guided by this analysis, three distinct strategies can be applied: retaining standard softmax attention in highly sensitive layers, replacing it with linear sliding window attention in moderately sensitive layers, and removing attention entirely in layers that exhibit low sensitivity.
To recover performance after these architectural modifications, we introduce a lightweight distillation-based healing phase requiring only 10M additional training tokens. LayerBoost reduces inference latency and improves throughput by up to 68% at high concurrency, while maintaining competitive model quality. It matches base model performance on several benchmarks, exhibits only minor degradations on others, and significantly outperforms state-of-the-art attention linearization methods. These efficiency gains make our method particularly well-suited for high-concurrency serving and hardware-constrained deployment scenarios, where inference cost and memory footprint are critical bottlenecks.
Fonte: arXiv cs.LG
Evaluation/Benchmarks • Score 85
Performance Anomaly Detection in Athletics: A Benchmarking System with Visual Analytics
arXiv:2604.21953v1 Announce Type: new
Abstract: Anti-doping programs rely on biological testing to detect performance-enhancing drugs, but such testing costs over $800 per sample and is limited by short detection windows for many prohibited substances. These constraints leave large portions of athletes without regular testing, motivating complementary screening approaches that analyze routine competition results to identify suspicious performance patterns. We present a system that processes 1.6 million athletics performances from over 19,000 competitions (2010-2025) using eight detection methods ranging from statistical rules to machine learning and trajectory analysis. We validate all methods against publicly confirmed anti-doping violations to measure their effectiveness in identifying sanctioned athletes. Trajectory-based methods, which compare performances to expected career progression, achieve the best balance between detecting violations and limiting false alarms, though all methods face challenges from incomplete data and rare confirmed violations. The system provides an interactive interface for expert-driven investigation, emphasizing transparency and human judgment to support, rather than replace, established anti-doping processes.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
QuantClaw: Precision Where It Matters for OpenClaw
arXiv:2604.22577v1 Announce Type: new
Abstract: Autonomous agent systems such as OpenClaw introduce significant efficiency challenges due to long-context inputs and multi-turn reasoning. This results in prohibitively high computational and monetary costs in real-world development. While quantization is a standard approach for reducing cost and latency, its impact on agent performance in realistic scenarios remains unclear. In this work, we analyze quantization sensitivity across diverse complex workflows over OpenClaw, and show that precision requirements are highly task-dependent. Based on this observation, we propose QuantClaw, a plug-and-play precision routing plugin that dynamically assigns precision according to task characteristics. QuantClaw routes lightweight tasks to lower-cost configurations while preserving higher precision for demanding workloads, saving cost and accelerating inference without increasing user complexity. Experiments show that our QuantClaw maintains or improves task performance while reducing both latency and computational cost. Across a range of agent tasks, it achieves up to 21.4% cost savings and 15.7% latency reduction on GLM-5 (FP8 baseline). These results highlight the benefit of treating precision as a dynamic resource in agent systems.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Call-Chain-Aware LLM-Based Test Generation for Java Projects
arXiv:2604.22046v1 Announce Type: cross
Abstract: Large language models (LLMs) have recently shown strong potential for generating project-level unit tests. However, existing state-of-the-art approaches primarily rely on execution-path information to guide prompt construction, which is often insufficient for complex software systems with rich inter-class dependencies, deep call chains, and intricate object initialization requirements. In this paper, we present CAT, a novel call-chain-aware LLM-based test generation approach that explicitly incorporates call-chain and dependency contexts into prompts through dedicated static analysis. To construct executable, semantically valid test contexts, CAT systematically models caller--callee relationships, object constructors, and third-party dependencies, and supports iterative test fixing when generation failures occur. We evaluate CAT on the widely used Defects4J benchmark and on four real-world GitHub projects released after the LLM's cut-off date. The results show that, across projects in Defects4J, CAT improves line and branch coverage by 18.04% and 21.74%, respectively, over the state-of-the-art approach PANTA, while consistently achieving superior performance on post-cutoff real-world projects. An ablation study further demonstrates the importance of call-chain and dependency contexts in CAT.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Generating Synthetic Malware Samples Using Generative AI
arXiv:2604.22084v1 Announce Type: new
Abstract: Malware attacks have a significant negative impact on organizations of varied scales in the field of cybersecurity. Recently, malware researchers have increasingly turned to machine learning techniques to combat sophisticated obfuscation methods used in malware. However, collecting a diverse set of malware samples with various obfuscation techniques is challenging and often takes years, especially for newly developed malware. This issue is further compounded by a well-known limitation of machine learning models: their poor performance when training data is scarce. In this paper, we propose a new system for generating synthetic malware samples to augment imbalanced malware dataset. Our approach decomposes malware binary samples into mnemonic opcode sequences, leveraging natural language processing to extract contextual meaning behind malware opcode features to aid the learning of generative AI (GenAI) employed in this paper, Generative Adversarial Networks (GAN), Wasserstein Generative Adversarial Networks with Gradient Penalty (WGAN-GP), and a modified Diffusion model. The experiment results show that augmenting training data with Diffusion-based synthetic data significantly improves classification performance for minor classes by up to 60% on average. This enhancement ultimately leads to an overall malware classification performance of 96%, an 8% improvement. These findings demonstrate the high quality and fidelity of the synthetic data, its robustness, and its potential applications in malware analysis. Specifically, synthetic malware data proves effective in improving the classification of minor malware classes and detection rates, even though the size of known malware data is significantly small.
Fonte: arXiv cs.LG
Applications • Score 90
MolClaw: An Autonomous Agent with Hierarchical Skills for Drug Molecule Evaluation, Screening, and Optimization
arXiv:2604.21937v1 Announce Type: new
Abstract: Computational drug discovery, particularly the complex workflows of drug molecule screening and optimization, requires orchestrating dozens of specialized tools in multi-step workflows, yet current AI agents struggle to maintain robust performance and consistently underperform in these high-complexity scenarios. Here we present MolClaw, an autonomous agent that leads drug molecule evaluation, screening, and optimization. It unifies over 30 specialized domain resources through a three-tier hierarchical skill architecture (70 skills in total) that facilitates agent long-term interaction at runtime: tool-level skills standardize atomic operations, workflow-level skills compose them into validated pipelines with quality check and reflection, and a discipline-level skill supplies scientific principles governing planning and verification across all scenarios in the field. Additionally, we introduce MolBench, a benchmark comprising molecular screening, optimization, and end-to-end discovery challenges spanning 8 to 50+ sequential tool calls. MolClaw achieves state-of-the-art performance across all metrics, and ablation studies confirm that gains concentrate on tasks that demand structured workflows while vanishing on those solvable with ad hoc scripting, establishing workflow orchestration competence as the primary capability bottleneck for AI-driven drug discovery.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
MambaCSP: Hybrid-Attention State Space Models for Hardware-Efficient Channel State Prediction
arXiv:2604.21957v1 Announce Type: cross
Abstract: Recent works have demonstrated that attention-based transformer and large language model (LLM) architectures can achieve strong channel state prediction (CSP) performance by capturing long-range temporal dependencies across channel state information (CSI) sequences. However, these models suffer from quadratic scaling in sequence length, leading to substantial computational cost, memory consumption, and inference latency, which limits their applicability in real-time and resource-constrained wireless deployments. In this paper, we investigate whether selective state space models (SSMs) can serve as a hardware-efficient alternative for CSI prediction. We propose MambaCSP, a hybrid-attention SSM architecture that replaces LLM-based prediction backbones with a linear-time Mamba model. To overcome the local-only dependencies of pure SSMs, we introduce lightweight patch-mixer attention layers that periodically inject cross-token attentions, helping with long-context CSI prediction. Extensive MISO-OFDM simulations show that MambaCSP improves prediction accuracy over LLM-based approaches by 9-12%, while delivering up to 3.0x higher throughput, 2.6x lower VRAM usage, and 2.9x faster inference. Our results demonstrate that hybrid state space architectures provide a promising direction for scalable and hardware-efficient AI-native CSI prediction in future wireless networks.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Removing Sandbagging in LLMs by Training with Weak Supervision
arXiv:2604.22082v1 Announce Type: cross
Abstract: As AI systems begin to automate complex tasks, supervision increasingly relies on weaker models or limited human oversight that cannot fully verify output quality. A model more capable than its supervisors could exploit this gap through sandbagging, producing work that appears acceptable but falls short of its true abilities. Can training elicit a model's best work even without reliable verification? We study this using model organisms trained to sandbag, testing elicitation techniques on problem-solving math, graduate-level science, and competitive coding tasks. We find that training with weak supervision can reliably elicit sandbagging models when supervised fine-tuning (SFT) and reinforcement learning (RL) are combined: SFT on weak demonstrations breaks the sandbagging behavior, enabling RL to then fully elicit performance. Neither method succeeds reliably alone-RL without SFT almost always leads to reward hacking rather than genuine improvement. Critically, this relies on training being indistinguishable from deployment; when models can distinguish between training and deployment, they can perform well during training while continuing to sandbag afterward. Our results provide initial evidence that training is a viable mitigation against sandbagging, while highlighting the importance of making training indistinguishable from deployment.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
An Artifact-based Agent Framework for Adaptive and Reproducible Medical Image Processing
arXiv:2604.21936v1 Announce Type: new
Abstract: Medical imaging research is increasingly shifting from controlled benchmark evaluation toward real-world clinical deployment. In such settings, applying analytical methods extends beyond model design to require dataset-aware workflow configuration and provenance tracking. Two requirements therefore become central: \textbf{adaptability}, the ability to configure workflows according to dataset-specific conditions and evolving analytical goals; and \textbf{reproducibility}, the guarantee that all transformations and decisions are explicitly recorded and re-executable. Here, we present an artifact-based agent framework that introduces a semantic layer to augment medical image processing. The framework formalizes intermediate and final outputs through an artifact contract, enabling structured interrogation of workflow state and goal-conditioned assembly of configurations from a modular rule library. Execution is delegated to a workflow executor to preserve deterministic computational graph construction and provenance tracking, while the agent operates locally to comply with most privacy constraints. We evaluate the framework on real-world clinical CT and MRI cohorts, demonstrating adaptive configuration synthesis, deterministic reproducibility across repeated executions, and artifact-grounded semantic querying. These results show that adaptive workflow configuration can be achieved without compromising reproducibility in heterogeneous clinical environments.
Fonte: arXiv cs.AI
MLOps/Systems • Score 92
Memanto: Typed Semantic Memory with Information-Theoretic Retrieval for Long-Horizon Agents
arXiv:2604.22085v1 Announce Type: new
Abstract: The transition from stateless language model inference to persistent, multi session autonomous agents has revealed memory to be a primary architectural bottleneck in the deployment of production grade agentic systems. Existing methodologies largely depend on hybrid semantic graph architectures, which impose substantial computational overhead during both ingestion and retrieval. These systems typically require large language model mediated entity extraction, explicit graph schema maintenance, and multi query retrieval pipelines. This paper introduces Memanto, a universal memory layer for agentic artificial intelligence that challenges the prevailing assumption that knowledge graph complexity is necessary to achieve high fidelity agent memory. Memanto integrates a typed semantic memory schema comprising thirteen predefined memory categories, an automated conflict resolution mechanism, and temporal versioning. These components are enabled by Moorcheh's Information Theoretic Search engine, a no indexing semantic database that provides deterministic retrieval within sub ninety millisecond latency while eliminating ingestion delay. Through systematic benchmarking on the LongMemEval and LoCoMo evaluation suites, Memanto achieves state of the art accuracy scores of 89.8 percent and 87.1 percent respectively. These results surpass all evaluated hybrid graph and vector based systems while requiring only a single retrieval query, incurring no ingestion cost, and maintaining substantially lower operational complexity. A five stage progressive ablation study is presented to quantify the contribution of each architectural component, followed by a discussion of the implications for scalable deployment of agentic memory systems.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Deep Learning for Model Calibration in Simulation of Itaconic Acid Production
arXiv:2604.22496v1 Announce Type: new
Abstract: In this study, deep learning is used to estimate kinetic parameters for modeling itaconic acid production based on real batch experiments conducted at different agitation speeds and reactor scales. Two deep learning strategies, namely direct deep learning (DDL) and generative conditional flow matching (CFM) are compared and benchmarked against nonlinear regression as a reference method. Compared with DDL, CFM consistently yields more accurate results. The concentration profiles predicted by CFM closely match those obtained from nonlinear regression, whereas DDL results in larger deviations. Similar behavior is observed in the scale-up experiments, where the CFM model again generalizes better and is more robust than the direct approach. These findings demonstrate that CFM can reliably predict system behavior across different operating conditions and scales, offering a flexible and data-efficient framework for parameter estimation in dynamic bioprocess models.
Fonte: arXiv cs.LG
RL • Score 85
SOLAR-RL: Semi-Online Long-horizon Assignment Reinforcement Learning
arXiv:2604.22558v1 Announce Type: new
Abstract: As Multimodal Large Language Models (MLLMs) mature, GUI agents are evolving from static interactions to complex navigation. While Reinforcement Learning (RL) has emerged as a promising paradigm for training MLLM agents on dynamic GUI tasks, its effective application faces a dilemma. Standard Offline RL often relies on static step-level data, neglecting global trajectory semantics such as task completion and execution quality. Conversely, Online RL captures the long-term dynamics but suffers from high interaction costs and potential environmental instability. To bridge this gap, we propose SOLAR-RL (Semi-Online Long-horizon Assignment Reinforcement Learning). Instead of relying solely on expensive online interactions, our framework integrates global trajectory insights directly into the offline learning process. Specifically, we reconstruct diverse rollout candidates from static data, detect the first failure point using per-step validity signals, and retroactively assign dense step-level rewards with target-aligned shaping to reflect trajectory-level execution quality, effectively simulating online feedback without interaction costs. Extensive experiments demonstrate that SOLAR-RL significantly improves long-horizon task completion rates and robustness compared to strong baselines, offering a sample-efficient solution for autonomous GUI navigation.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Shard the Gradient, Scale the Model: Serverless Federated Aggregation via Gradient Partitioning
arXiv:2604.22072v1 Announce Type: cross
Abstract: Federated learning (FL) aggregation on serverless platforms faces a hard scalability ceiling: existing architectures (lambda-FL, LIFL) partition clients across aggregators, but every aggregator must hold the complete model gradient in memory. When gradients exceed the per-function memory limit (e.g., 10 GB on AWS Lambda), aggregation becomes infeasible regardless of tree depth or branching factor. We propose GradsSharding, which instead partitions the gradient tensor into M shards, each averaged independently by a serverless function that receives contributions from all clients. Because FedAvg averaging is element-wise, this produces bit-identical results to tree-based approaches, so model accuracy is invariant by construction. Per-function memory is bounded at O(|{\theta}|/M), independent of client count, enabling aggregation of arbitrarily large models. We evaluate GradsSharding against lambda-FL and LIFL through HPC experiments and real AWS Lambda deployments across model sizes from 43 MB to 5 GB. Results show a cost crossover at approximately 500 MB gradient size, 2.7x cost reduction at VGG-16 scale, and that GradsSharding is the only architecture that remains deployable beyond the serverless memory ceiling.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Kernel Contracts: A Specification Language for ML Kernel Correctness Across Heterogeneous Silicon
arXiv:2604.22032v1 Announce Type: new
Abstract: Every ML kernel ships with an implicit contract about what it computes. People rarely write the contract down. When two kernels disagree -- when a matmul on AMD produces a different gradient than the same matmul on NVIDIA, when a fused attention kernel silently downcasts an accumulator, when an out-of-bounds access returns zero on one stack and garbage on another -- there is no formal artifact to arbitrate the dispute. Recent empirical work has measured the gap across silicon platforms, but none of it specifies the contract being violated.
We present a specification language for kernel contracts. A contract has eight parts: identifier, scope, precondition, postcondition, tolerance, reference oracle, measurement protocol, and violation signature. We use it to state twelve contract classes covering precision, ordering, compiler-induced, and exceptional-value failure modes, each grounded in published empirical evidence. We require a three-state calibration: every contract must admit at least one reference-conforming implementation and at least one contract-violating implementation that passes basic functional tests. We apply the framework to three documented incidents -- Huawei Ascend silent precision coercion, Sakana AI CUDA Engineer reward hacking, AMD out-of-bounds silent acceptance -- and show that each informal diagnosis maps to a specific contract violation with a measurable signature. A kernel contract suite is a normative reference against which conformance can be graded, in the way that ISASecure grades industrial control systems against IEC 62443.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Towards Adaptive Continual Model Merging via Manifold-Aware Expert Evolution
arXiv:2604.22464v1 Announce Type: new
Abstract: Continual Model Merging (CMM) sequentially integrates task-specific models into a unified architecture without intensive retraining. However, existing CMM methods are hindered by a fundamental saturation-redundancy dilemma: backbone-centric approaches face parameter saturation and representation interference within fixed capacities, whereas Mixture-of-Experts (MoE) variants resort to indiscriminate expansion, incurring expert redundancy and a routing bottleneck reliant on additional data-driven optimization. To resolve these challenges, we propose MADE-IT (Manifold-Aware Dynamic Expert Evolution and Implicit rouTing), an adaptive CMM method that orchestrates expert management and activation by grounding intrinsic expert representations in manifold geometry. We introduce a projection-based subspace affinity metric coupled with a distribution-aware adaptive threshold mechanism to guide autonomous expert evolution, harmonizing diversity with architectural parsimony. Furthermore, to bypass parameterized gating networks, we design a data-free and training-free implicit routing mechanism that activates experts via feature-subspace alignment. Extensive experiments demonstrate that MADE-IT consistently outperforms strong baselines in accuracy and robustness across long-horizon and shuffled task sequences, while significantly pruning redundant experts, particularly within generic modules and early layers.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Hidden Failure Modes of Gradient Modification under Adam in Continual Learning, and Adaptive Decoupled Moment Routing as a Repair
arXiv:2604.22407v1 Announce Type: new
Abstract: Many continual-learning methods modify gradients upstream (e.g., projection, penalty rescaling, replay mixing) while treating Adam as a neutral backend. We show this composition has a hidden failure mode. In a high-overlap, non-adaptive 8-domain continual LM, all shared-routing projection baselines collapse close to vanilla forgetting (12.5--12.8 vs. 13.2). A 0.5% replay buffer is the strongest shared alternative but still reaches 11.6, while fixed-strength decoupling falls below vanilla at 14.1. Only adaptive decoupled routing remains stable at 9.4, improving over vanilla by 3.8 units. On a 16-domain stream, its gain over the strongest shared-routing projection baseline grows to 4.5--4.8 units. The failure is largely invisible on clean benchmarks.
We explain this effect through Adam's second-moment pathway: in the tested regime, projection induces a 1/(1-alpha) inflation of the old-direction effective learning rate, matching measurements within 8% across eight alpha values. The same conflict appears with penalty methods, replay mixing, and at 7B scale under LoRA. Our fix routes the modified gradient only to the first moment while preserving magnitude-faithful second-moment statistics, with overlap-aware adaptive strength. This simple change is the only tested configuration that consistently avoids collapse across methods, optimizers, and scale.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
SOC-ICNN: From Polyhedral to Conic Geometry for Learning Convex Surrogate Functions
arXiv:2604.22355v1 Announce Type: new
Abstract: Classical ReLU-based Input Convex Neural Networks (ICNNs) are equivalent to the optimal value functions of Linear Programming (LP). This intrinsic structural equivalence restricts their representational capacity to piecewise-linear polyhedral functions. To overcome this representational bottleneck, we propose the SOC-ICNN, an architecture that generalizes the underlying optimization class from LP to Second-Order Cone Programming (SOCP). By explicitly injecting positive semi-definite curvature and Euclidean norm-based conic primitives, our formulation introduces native smooth curvature into the representation while preserving a rigorous optimization-theoretic interpretation. We formally prove that SOC-ICNNs strictly expand the representational space of ReLU-ICNNs without increasing the asymptotic order of forward-pass complexity. Extensive experiments demonstrate that SOC-ICNN substantially improves function approximation, while delivering competitive downstream decision quality. The code is available at https://github.com/Kanyooo/SOC-ICNN.
Fonte: arXiv cs.LG
Applications • Score 85
Architecture of an AI-Based Automated Course of Action Generation System for Military Operations
arXiv:2604.20862v1 Announce Type: new
Abstract: The automation system for Course of Action (CoA) planning is an essential element in future warfare. As maneuver speeds increase, surveillance ranges extend, and weapon ranges grow, the operational area expands, making traditional manned-based CoA planning increasingly challenging. Consequently, the development of an AI-based automated CoA planning system is becoming increasingly necessary. Accordingly, several countries and defense organizations are actively developing AI-based CoA planning systems. However, due to security restrictions and limited public disclosure, the technical maturity of such systems remains difficult to assess. Furthermore, as these systems are military-related, their details are not publicly disclosed, making it difficult to accurately assess the current level of development. In response to this, this study aims to introduce relevant doctrines within the scope of publicly available information and present applicable AI technologies for each stage of the CoA planning process. Ultimately, it proposes an architecture for the development of an automated CoA planning system.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
ImageHD: Energy-Efficient On-Device Continual Learning of Visual Representations via Hyperdimensional Computing
arXiv:2604.21280v1 Announce Type: new
Abstract: On-device continual learning (CL) is critical for edge AI systems operating on non-stationary data streams, but most existing methods rely on backpropagation or exemplar-heavy classifiers, incurring substantial compute, memory, and latency overheads. Hyperdimensional computing (HDC) offers a lightweight alternative through fast, non-iterative online updates. Combined with a compact convolutional neural network (CNN) feature extractor, HDC enables efficient on-device adaptation with strong visual representations. However, prior HDC-based CL systems often depend on multi-tier memory hierarchies and complex cluster management, limiting deployability on resource-constrained hardware.
We present ImageHD, an FPGA accelerator for on-device continual learning of visual data based on HDC. ImageHD targets streaming CL under strict latency and on-chip memory constraints, avoiding costly iterative optimization. At the algorithmic level, we introduce a hardware-aware CL method that bounds class exemplars through a unified exemplar memory and a hardware-efficient cluster merging strategy, while incorporating a quantized CNN front-end to reduce deployment overhead without sacrificing accuracy. At the system level, ImageHD is implemented as a streaming dataflow architecture on the AMD Zynq ZCU104 FPGA, integrating HDC encoding, similarity search, and bounded cluster management using word-packed binary hypervectors for massively parallel bitwise computation within tight on-chip resource budgets. On CORe50, ImageHD achieves up to 40.4x (4.84x) speedup and 383x (105.1x) energy efficiency over optimized CPU (GPU) baselines, demonstrating the practicality of HDC-enabled continual learning for real-time edge AI.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
Decoupled DiLoCo for Resilient Distributed Pre-training
arXiv:2604.21428v1 Announce Type: new
Abstract: Modern large-scale language model pre-training relies heavily on the single program multiple data (SPMD) paradigm, which requires tight coupling across accelerators. Due to this coupling, transient slowdowns, hardware failures, and synchronization overhead stall the entire computation, wasting significant compute time at scale. While recent distributed methods like DiLoCo reduced communication bandwidth, they remained fundamentally synchronous and vulnerable to these system stalls. To address this, we introduce Decoupled DiLoCo, an evolution of the DiLoCo framework designed to break the lock-step synchronization barrier and go beyond SPMD to maximize training goodput. Decoupled DiLoCo partitions compute across multiple independent ``learners'' that execute local inner optimization steps. These learners asynchronously communicate parameter fragments to a central synchronizer, which circumvents failed or straggling learners by aggregating updates using a minimum quorum, an adaptive grace window, and dynamic token-weighted merging. Inspired by ``chaos engineering'', we achieve significantly improved training efficiency in failure-prone environments with millions of simulated chips with strictly zero global downtime, while maintaining competitive model performance across text and vision tasks, for both dense and mixture-of-expert architectures.
Fonte: arXiv cs.CL
Vision • Score 85
HyperFM: An Efficient Hyperspectral Foundation Model with Spectral Grouping
arXiv:2604.21127v1 Announce Type: new
Abstract: The NASA PACE mission provides unprecedented hyperspectral observations of ocean color, aerosols, and clouds, offering new insights into how these components interact and influence Earth's climate and air quality. Its Ocean Color Instrument measures light across hundreds of finely spaced wavelength bands, enabling detailed characterization of features such as phytoplankton composition, aerosol properties, and cloud microphysics. However, hyperspectral data of this scale is large, complex, and difficult to label, requiring specialized processing and analysis techniques. Existing foundation models, which have transformed computer vision and natural language processing, are generally trained on standard RGB imagery and therefore struggle to interpret the continuous spectral signatures captured by PACE. While recent advances have introduced hyperspectral foundation models, they are typically trained on cloud-free observations and often remain limited to single-sensor datasets due to spectral inconsistencies across instruments. Moreover, existing models tend to be parameter-heavy and computationally expensive, limiting scalability and adoption in operational settings. To address these challenges, we introduce HyperFM, a parameter-efficient hyperspectral foundation model that leverages intra-group and inter-group spectral attention along with hybrid parameter decomposition to better capture spectral spatial relationships while reducing computational cost. HyperFM demonstrates consistent performance improvements over existing hyperspectral foundation models and task-specific state-of-the-art methods across four benchmark downstream atmospheric cloud property retrieval tasks. To support further research, we additionally release HyperFM250K, a large-scale hyperspectral dataset from the PACE mission that includes both clear and cloudy scenes.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Beyond N-gram: Data-Aware X-GRAM Extraction for Efficient Embedding Parameter Scaling
arXiv:2604.21724v1 Announce Type: new
Abstract: Large token-indexed lookup tables provide a compute-decoupled scaling path, but their practical gains are often limited by poor parameter efficiency and rapid memory growth. We attribute these limitations to Zipfian under-training of the long tail, heterogeneous demand across layers, and "slot collapse" that produces redundant embeddings. To address this, we propose X-GRAM, a frequency-aware dynamic token-injection framework. X-GRAM employs hybrid hashing and alias mixing to compress the tail while preserving head capacity, and refines retrieved vectors via normalized SwiGLU ShortConv to extract diverse local n-gram features. These signals are integrated into attention value streams and inter-layer residuals using depth-aware gating, effectively aligning static memory with dynamic context. This design introduces a memory-centric scaling axis that decouples model capacity from FLOPs. Extensive evaluations at the 0.73B and 1.15B scales show that X-GRAM improves average accuracy by as much as 4.4 points over the vanilla backbone and 3.2 points over strong retrieval baselines, while using substantially smaller tables in the 50% configuration. Overall, by decoupling capacity from compute through efficient memory management, X-GRAM offers a scalable and practical paradigm for future memory-augmented architectures. Code aviliable in https://github.com/Longyichen/X-gram.
Fonte: arXiv cs.CL
Vision • Score 85
GraphLeap: Decoupling Graph Construction and Convolution for Vision GNN Acceleration on FPGA
arXiv:2604.21290v1 Announce Type: new
Abstract: Vision Graph Neural Networks (ViGs) represent an image as a graph of patch tokens, enabling adaptive, feature-driven neighborhoods. Unlike CNNs with fixed grid biases or Vision Transformers with global token interactions, ViGs rely on dynamic graph convolution: at each layer, a feature-dependent graph is built via k-nearest-neighbor (kNN) search on current patch features, followed by message passing. This per-layer graph construction is the main bottleneck, consuming 50--95\% of graph convolution time on CPUs and GPUs, scaling as $O(N^2)$ with the number of patches $N$, and creating a sequential dependency between graph construction and feature updates.
We introduce GraphLeap, a simple reformulation that removes this dependency by decoupling graph construction from feature update across layers. GraphLeap performs the feature update at layer $\ell$ using a graph built from the previous layer's features, while simultaneously using the current layer's features to construct the graph for layer $\ell+1$. This one-layer-lookahead graph construction enables concurrent graph construction and message passing. Although using prior-layer features can introduce minor accuracy degradation, lightweight fine-tuning for a few epochs is sufficient to recover the original accuracy. Building on GraphLeap, we present the first end-to-end FPGA accelerator for Vision GNNs. Our streaming, layer-pipelined design overlaps a kNN graph construction engine with a feature update engine, exploits node- and channel-level parallelism, and enables efficient on-chip dataflow without explicit edge-feature materialization. Evaluated on isotropic and pyramidal ViG models on an Alveo U280 FPGA, GraphLeap achieves up to $95.7\times$ speedup over CPU and $8.5\times$ speedup over GPU baselines, demonstrating the feasibility of real-time Vision GNN inference.
Fonte: arXiv cs.CV
RL • Score 85
AEL: Agent Evolving Learning for Open-Ended Environments
arXiv:2604.21725v1 Announce Type: new
Abstract: LLM agents increasingly operate in open-ended environments spanning hundreds of sequential episodes, yet they remain largely stateless: each task is solved from scratch without converting past experience into better future behavior. The central obstacle is not \emph{what} to remember but \emph{how to use} what has been remembered, including which retrieval policy to apply, how to interpret prior outcomes, and when the current strategy itself must change. We introduce \emph{Agent Evolving Learning} (\ael{}), a two-timescale framework that addresses this obstacle. At the fast timescale, a Thompson Sampling bandit learns which memory retrieval policy to apply at each episode; at the slow timescale, LLM-driven reflection diagnoses failure patterns and injects causal insights into the agent's decision prompt, giving it an interpretive frame for the evidence it retrieves. On a sequential portfolio benchmark (10 sector-diverse tickers, 208 episodes, 5 random seeds), \ael{} achieves a Sharpe ratio of 2.13$\pm$0.47, outperforming five published self-improving methods and all non-LLM baselines while maintaining the lowest variance among all LLM-based approaches. A nine-variant ablation reveals a ``less is more'' pattern: memory and reflection together produce a 58\% cumulative improvement over the stateless baseline, yet every additional mechanism we test (planner evolution, per-tool selection, cold-start initialization, skill extraction, and three credit assignment methods) \emph{degrades} performance. This demonstrates that the bottleneck in agent self-improvement is \emph{self-diagnosing how to use} experience rather than adding architectural complexity. Code and data: https://github.com/WujiangXu/AEL.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
AgenticQwen: Training Small Agentic Language Models with Dual Data Flywheels for Industrial-Scale Tool Use
arXiv:2604.21590v1 Announce Type: new
Abstract: Modern industrial applications increasingly demand language models that act as agents, capable of multi-step reasoning and tool use in real-world settings. These tasks are typically performed under strict cost and latency constraints, making small agentic models highly desirable. In this paper, we introduce the AgenticQwen family of models, trained via multi-round reinforcement learning (RL) on synthetic data and a limited amount of open-source data. Our training framework combines reasoning RL and agentic RL with dual data flywheels that automatically generate increasingly challenging tasks. The reasoning flywheel increases task difficulty by learning from errors, while the agentic flywheel expands linear workflows into multi-branch behavior trees that better reflect the decision complexity of real-world applications. We validate AgenticQwen on public benchmarks and in an industrial agent system. The models achieve strong performance on multiple agentic benchmarks, and in our industrial agent system, close the gap with much larger models on search and data analysis tasks. Model checkpoints and part of the synthetic data: https://huggingface.co/collections/alibaba-pai/agenticqwen. Data synthesis and RL training code: https://github.com/haruhi-sudo/data_synth_and_rl. The data synthesis pipeline is also integrated into EasyDistill: https://github.com/modelscope/easydistill.
Fonte: arXiv cs.CL
Privacy/Security/Fairness • Score 85
Differentially Private Model Merging
arXiv:2604.20985v1 Announce Type: cross
Abstract: In machine learning applications, privacy requirements during inference or deployment time could change constantly due to varying policies, regulations, or user experience. In this work, we aim to generate a magnitude of models to satisfy any target differential privacy (DP) requirement without additional training steps, given a set of existing models trained on the same dataset with different privacy/utility tradeoffs. We propose two post processing techniques, namely random selection and linear combination, to output a final private model for any target privacy parameter. We provide privacy accounting of these approaches from the lens of R'enyi DP and privacy loss distributions for general problems. In a case study on private mean estimation, we fully characterize the privacy/utility results and theoretically establish the superiority of linear combination over random selection. Empirically, we validate our approach and analyses on several models and both synthetic and real-world datasets.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Projected Gradient Unlearning for Text-to-Image Diffusion Models: Defending Against Concept Revival Attacks
arXiv:2604.21041v1 Announce Type: new
Abstract: Machine unlearning for text-to-image diffusion models aims to selectively remove undesirable concepts from pre-trained models without costly retraining. Current unlearning methods share a common weakness: erased concepts return when the model is fine-tuned on downstream data, even when that data is entirely unrelated. We adapt Projected Gradient Unlearning (PGU) from classification to the diffusion domain as a post-hoc hardening step. By constructing a Core Gradient Space (CGS) from the retain concept activations and projecting gradient updates into its orthogonal complement, PGU ensures that subsequent fine-tuning cannot undo the achieved erasure. Applied on top of existing methods (ESD, UCE, Receler), the approach eliminates revival for style concepts and substantially delays it for object concepts, running in roughly 6 minutes versus the ~2 hours required by Meta-Unlearning. PGU and Meta-Unlearning turn out to be complementary: which performs better depends on how the concept is encoded, and retain concept selection should follow visual feature similarity rather than semantic grouping.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Adaptive Conformal Anomaly Detection with Time Series Foundation Models for Signal Monitoring
arXiv:2604.20122v1 Announce Type: new
Abstract: We propose a post-hoc adaptive conformal anomaly detection method for monitoring time series that leverages predictions from pre-trained foundation models without requiring additional fine-tuning. Our method yields an interpretable anomaly score directly interpretable as a false alarm rate (p-value), facilitating transparent and actionable decision-making. It employs weighted quantile conformal prediction bounds and adaptively learns optimal weighting parameters from past predictions, enabling calibration under distribution shifts and stable false alarm control, while preserving out-of-sample guarantees. As a model-agnostic solution, it integrates seamlessly with foundation models and supports rapid deployment in resource-constrained environments. This approach addresses key industrial challenges such as limited data availability, lack of training expertise, and the need for immediate inference, while taking advantage of the growing accessibility of time series foundation models. Experiments on both synthetic and real-world datasets show that the proposed approach delivers strong performance, combining simplicity, interpretability, robustness, and adaptivity.
Fonte: arXiv cs.LG
Vision • Score 85
Prototype-Based Test-Time Adaptation of Vision-Language Models
arXiv:2604.21360v1 Announce Type: new
Abstract: Test-time adaptation (TTA) has emerged as a promising paradigm for vision-language models (VLMs) to bridge the distribution gap between pre-training and test data. Recent works have focused on backpropagation-free TTA methods that rely on cache-based designs, but these introduce two key limitations. First, inference latency increases as the cache grows with the number of classes, leading to inefficiencies in large-scale settings. Second, suboptimal performance occurs when the cache contains insufficient or incorrect samples. In this paper, we present Prototype-Based Test-Time Adaptation (PTA), an efficient and effective TTA paradigm that uses a set of class-specific knowledge prototypes to accumulate knowledge from test samples. Particularly, knowledge prototypes are adaptively weighted based on the zero-shot class confidence of each test sample, incorporating the sample's visual features into the corresponding class-specific prototype. It is worth highlighting that the knowledge from past test samples is integrated and utilized solely in the prototypes, eliminating the overhead of cache population and retrieval that hinders the efficiency of existing TTA methods. This endows PTA with extremely high efficiency while achieving state-of-the-art performance on 15 image recognition benchmarks and 4 robust point cloud analysis benchmarks. For example, PTA improves CLIP's accuracy from 65.64% to 69.38% on 10 cross-domain benchmarks, while retaining 92% of CLIP's inference speed on large-scale ImageNet-1K. In contrast, the cache-based TDA achieves a lower accuracy of 67.97% and operates at only 50% of CLIP's inference speed.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
On-Meter Graph Machine Learning: A Case Study of PV Power Forecasting for Grid Edge Intelligence
arXiv:2604.19800v1 Announce Type: new
Abstract: This paper presents a detailed study of how graph neural networks can be used on edge intelligent meters in a microgrid to forecast photovoltaic power generation. The problem background and the adopted technologies are introduced, including ONNX and ONNX Runtime. The hardware and software specifications of the smart meter are also briefly described. Then, the paper focuses on the training and deployment of two graph machine learning models, GCN and GraphSAGE, with particular emphasis on developing and deploying a customized ONNX operator for GCN. Finally, a case study is conducted using real datasets from a village microgrid. The performance of the two models is compared on both the PC and the smart meter, exhibiting successful deployments and executions on the smart meter.
Fonte: arXiv cs.LG
Vision • Score 85
You Only Gaussian Once: Controllable 3D Gaussian Splatting for Ultra-Densely Sampled Scenes
arXiv:2604.21400v1 Announce Type: new
Abstract: 3D Gaussian Splatting (3DGS) has revolutionized neural rendering, yet existing methods remain predominantly research prototypes ill-suited for production-level deployment. We identify a critical "Industry-Academia Gap" hindering real-world application: unpredictable resource consumption from heuristic Gaussian growth, the "sparsity shield" of current benchmarks that rewards hallucination over physical fidelity, and severe multi-sensor data pollution. To bridge this gap, we propose YOGO (You Only Gaussian Once), a system-level framework that reformulates the stochastic growth process into a deterministic, budget-aware equilibrium. YOGO integrates a novel budget controller for hardware-constrained resource allocation and an availability-registration protocol for robust multi-sensor fusion. To push the boundaries of reconstruction fidelity, we introduce Immersion v1.0, the first ultra-dense indoor dataset specifically designed to break the "sparsity shield." By providing saturated viewpoint coverage, Immersion v1.0 forces algorithms to focus on extreme physical fidelity rather than viewpoint interpolation, and enables the community to focus on the upper limits of high-fidelity reconstruction. Extensive experiments demonstrate that YOGO achieves state-of-the-art visual quality while maintaining a strictly deterministic profile, establishing a new standard for production-grade 3DGS. To facilitate reproducibility, part scenes of Immersion v1.0 dataset and source code of YOGO has been publicly released. The project link is https://jjrcn.github.io/YOGO/.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Accelerating PayPal's Commerce Agent with Speculative Decoding: An Empirical Study on EAGLE3 with Fine-Tuned Nemotron Models
arXiv:2604.19767v1 Announce Type: new
Abstract: We evaluate speculative decoding with EAGLE3 as an inference-time optimization for PayPal's Commerce Agent, powered by a fine-tuned llama3.1-nemotron-nano-8B-v1 model. Building on prior work (NEMO-4-PAYPAL) that reduced latency and cost through domain-specific fine-tuning, we benchmark EAGLE3 via vLLM against NVIDIA NIM on identical 2xH100 hardware across 40 configurations spanning speculative token counts (gamma=3, gamma=5), concurrency levels (1-32), and sampling temperatures (0, 0.5). Key findings: (1) gamma=3 achieves 22-49% throughput improvement and 18-33% latency reduction at zero additional hardware cost; (2) acceptance rates remain stable at approximately 35.5% for gamma=3 across all conditions; (3) gamma=5 yields diminishing returns (approximately 25% acceptance rate); (4) LLM-as-Judge evaluation confirms fully preserved output quality; and (5) speculative decoding on a single H100 matches or exceeds NIM on two H100s, enabling 50% GPU cost reduction.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Mind the Prompt: Self-adaptive Generation of Task Plan Explanations via LLMs
arXiv:2604.21092v1 Announce Type: new
Abstract: Integrating Large Language Models (LLMs) into complex software systems enables the generation of human-understandable explanations of opaque AI processes, such as automated task planning. However, the quality and reliability of these explanations heavily depend on effective prompt engineering. The lack of a systematic understanding of how diverse stakeholder groups formulate and refine prompts hinders the development of tools that can automate this process. We introduce COMPASS (COgnitive Modelling for Prompt Automated SynthesiS), a proof-of-concept self-adaptive approach that formalises prompt engineering as a cognitive and probabilistic decision-making process. COMPASS models unobservable users' latent cognitive states, such as attention and comprehension, uncertainty, and observable interaction cues as a POMDP, whose synthesised policy enables adaptive generation of explanations and prompt refinements. We evaluate COMPASS using two diverse cyber-physical system case studies to assess the adaptive explanation generation and their qualities, both quantitatively and qualitatively. Our results demonstrate the feasibility of COMPASS integrating human cognition and user profile's feedback into automated prompt synthesis in complex task planning systems.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Temporally Extended Mixture-of-Experts Models
arXiv:2604.20156v1 Announce Type: new
Abstract: Mixture-of-Experts models, now popular for scaling capacity at fixed inference speed, switch experts at nearly every token. Once a model outgrows available GPU memory, this churn can render optimizations like offloading and pre-fetching ineffective. We make the case that the options framework in reinforcement learning is a perfect match to tackle this problem, and argue for temporally extended mixture-of-experts layers. Building on the option-critic framework with deliberation costs, we add a controller to each layer that learns when to switch expert sets and which to load. By applying this to gpt-oss-20b with low-rank adapters and a self-distillation reward, our method reduces switch rates from over 50% to below 5% while retaining up to 90% of base-model accuracy on MATH, MMLU, and MMMLU. This shows that even existing pre-trained models can be converted to temporally extended MoEs with lightweight training, with the deliberation cost allowing model trainers to trade off switching rates against capability. We hope this opens a principled path, grounded in the options framework, for memory-efficient serving and continual learning in ever-growing MoE models.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
The Last Harness You'll Ever Build
arXiv:2604.21003v1 Announce Type: new
Abstract: AI agents are increasingly deployed on complex, domain-specific workflows -- navigating enterprise web applications that require dozens of clicks and form fills, orchestrating multi-step research pipelines that span search, extraction, and synthesis, automating code review across unfamiliar repositories, and handling customer escalations that demand nuanced domain knowledge. \textbf{Each new task domain requires painstaking, expert-driven harness engineering}: designing the prompts, tools, orchestration logic, and evaluation criteria that make a foundation model effective. We present a two-level framework that automates this process. At the first level, the \textbf{Harness Evolution Loop} optimizes a worker agent's harness $\mathcal{H}$ for a single task: a Worker Agent $W_{\mathcal{H}}$ executes the task, an Evaluator Agent $V$ adversarially diagnoses failures and scores performance, and an Evolution Agent $E$ modifies the harness based on the full history of prior attempts. At the second level, the \textbf{Meta-Evolution Loop} optimizes the evolution protocol $\Lambda = (W_{\mathcal{H}}, \mathcal{H}^{(0)}, V, E)$ itself across diverse tasks, \textbf{learning a protocol $\Lambda^{(\text{best})}$ that enables rapid harness convergence on any new task -- so that adapting an agent to a novel domain requires no human harness engineering at all.} We formalize the correspondence to meta-learning and present both algorithms. The framework \textbf{shifts manual harness engineering into automated harness engineering}, and takes one step further -- \textbf{automating the design of the automation itself}.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Generative Augmentation of Imbalanced Flight Records for Flight Diversion Prediction: A Multi-objective Optimisation Framework
arXiv:2604.20288v1 Announce Type: new
Abstract: Flight diversions are rare but high-impact events in aviation, making their reliable prediction vital for both safety and operational efficiency. However, their scarcity in historical records impedes the training of machine learning models utilised to predict them. This study addresses this scarcity gap by investigating how generative models can augment historical flight data with synthetic diversion records to enhance model training and improve predictive accuracy. We propose a multi-objective optimisation framework coupled with automated hyperparameter search to identify optimal configurations for three deep generative models: Tabular Variational Autoencoder (TVAE), Conditional Tabular Generative Adversarial Network (CTGAN), and CopulaGAN, with the Gaussian Copula (GC) model serving as a statistical baseline. The quality of the synthetic data was examined through a six-stage evaluation framework encompassing realism, diversity, operational validity, statistical similarity, fidelity, and predictive utility. Results show that the optimised models significantly outperform their non-optimised counterparts, and that synthetic augmentation substantially improves diversion prediction compared to models trained solely on real data. These findings demonstrate the effectiveness of hyperparameter-optimised generative models for advancing predictive modelling of rare events in air transportation.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
PCA-Driven Adaptive Sensor Triage for Edge AI Inference
arXiv:2604.05045v1 Announce Type: cross
Abstract: Multi-channel sensor networks in industrial IoT often exceed available bandwidth. We propose PCA-Triage, a streaming algorithm that converts incremental PCA loadings into proportional per-channel sampling rates under a bandwidth budget. PCA-Triage runs in O(wdk) time with zero trainable parameters (0.67 ms per decision).
We evaluate on 7 benchmarks (8--82 channels) against 9 baselines. PCA-Triage is the best unsupervised method on 3 of 6 datasets at 50% bandwidth, winning 5 of 6 against every baseline with large effect sizes (r = 0.71--0.91). On TEP, it achieves F1 = 0.961 +/- 0.001 -- within 0.1% of full-data performance -- while maintaining F1 > 0.90 at 30% budget. Targeted extensions push F1 to 0.970. The algorithm is robust to packet loss and sensor noise (3.7--4.8% degradation under combined worst-case).
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Improved large-scale graph learning through ridge spectral sparsification
arXiv:2604.20078v1 Announce Type: new
Abstract: Graph-based techniques and spectral graph theory have enriched the field of machine learning with a variety of critical advances. A central object in the analysis is the graph Laplacian L, which encodes the structure of the graph. We consider the problem of learning over this Laplacian in a distributed streaming setting, where new edges of the graph are observed in real time by a network of workers. In this setting, it is hard to learn quickly or approximately while keeping a distributed representation of L. To address this challenge, we present a novel algorithm, GSQUEAK, which efficiently sparsifies the Laplacian by maintaining a small subset of effective resistances. We show that our algorithm produces sparsifiers with strong spectral approximation guarantees, all while processing edges in a single pass and in a distributed fashion.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Synthetic Flight Data Generation Using Generative Models
arXiv:2604.20293v1 Announce Type: new
Abstract: The increasing adoption of synthetic data in aviation research offers a promising solution to data scarcity and confidentiality challenges. This study investigates the potential of generative models to produce realistic synthetic flight data and evaluates their quality through a comprehensive four-stage assessment framework. The need for synthetic flight data arises from their potential to serve as an alternative to confidential real-world records and to augment rare events in historical datasets. These enhanced datasets can then be used to train machine learning models that predict critical events, such as flight delays, cancellations, diversions, and turnaround times. Two generative models, Tabular Variational Autoencoder (TVAE) and Gaussian Copula (GC), are adapted to generate synthetic flight information and compared based on their ability to preserve statistical similarity, fidelity, diversity, and predictive utility. Results indicate that while GC achieves higher statistical similarity and fidelity, its computational cost hinders its applicability to large datasets. In contrast, TVAE efficiently handles large datasets and enables scalable synthetic data generation. The findings demonstrate that synthetic data can support flight delay prediction models with accuracy comparable to those trained on real data. These results pave the way for leveraging synthetic flight data to enhance predictive modeling in air transportation.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
WorkflowGen:an adaptive workflow generation mechanism driven by trajectory experience
arXiv:2604.19756v1 Announce Type: new
Abstract: Large language model (LLM) agents often suffer from high reasoning overhead, excessive token consumption, unstable execution, and inability to reuse past experiences in complex tasks like business queries, tool use, and workflow orchestration. Traditional methods generate workflows from scratch for every query, leading to high cost, slow response, and poor robustness. We propose WorkflowGen, an adaptive, trajectory experience-driven framework for automatic workflow generation that reduces token usage and improves efficiency and success rate. Early in execution, WorkflowGen captures full trajectories and extracts reusable knowledge at both node and workflow levels, including error fingerprints, optimal tool mappings, parameter schemas, execution paths, and exception-avoidance strategies. It then employs a closed-loop mechanism that performs lightweight generation only on variable nodes via trajectory rewriting, experience updating, and template induction. A three-tier adaptive routing strategy dynamically selects among direct reuse, rewriting-based generation, and full initialization based on semantic similarity to historical queries. Without large annotated datasets, we qualitatively compare WorkflowGen against real-time planning, static single trajectory, and basic in-context learning baselines. Our method reduces token consumption by over 40 percent compared to real-time planning, improves success rate by 20 percent on medium-similarity queries through proactive error avoidance and adaptive fallback, and enhances deployability via modular, traceable experiences and cross-scenario adaptability. WorkflowGen achieves a practical balance of efficiency, robustness, and interpretability, addressing key limitations of existing approaches.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Spatial Metaphors for LLM Memory: A Critical Analysis of the MemPalace Architecture
arXiv:2604.21284v1 Announce Type: new
Abstract: MemPalace is an open-source AI memory system that applies the ancient method of loci (memory palace) spatial metaphor to organize long-term memory for large language models; launched in April 2026, it accumulated over 47,000 GitHub stars in its first two weeks and claims state-of-the-art retrieval performance on the LongMemEval benchmark (96.6% Recall@5) without requiring any LLM inference at write time. Through independent codebase analysis, benchmark replication, and comparison with competing systems, we find that MemPalace's headline retrieval performance is attributable primarily to its verbatim storage philosophy combined with ChromaDB's default embedding model (all-MiniLM-L6-v2), rather than to its spatial organizational metaphor per se -- the palace hierarchy (Wings->Rooms->Closets->Drawers) operates as standard vector database metadata filtering, an effective but well-established technique. However, MemPalace makes several genuinely novel contributions: (1) a contrarian verbatim-first storage philosophy that challenges extraction-based competitors, (2) an extremely low wake-up cost (approximately 170 tokens) through its four-layer memory stack, (3) a fully deterministic, zero-LLM write path enabling offline operation at zero API cost, and (4) the first systematic application of spatial memory metaphors as an organizing principle for AI memory systems. We also note that the competitive landscape is evolving rapidly, with Mem0's April 2026 token-efficient algorithm raising their LongMemEval score from approximately 49% to 93.4%, narrowing the gap between extraction-based and verbatim approaches. Our analysis concludes that MemPalace represents significant architectural insight wrapped in overstated claims -- a pattern common in rapidly adopted open-source projects where marketing velocity exceeds scientific rigor.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Time, Causality, and Observability Failures in Distributed AI Inference Systems
arXiv:2604.21361v1 Announce Type: new
Abstract: Distributed AI inference pipelines rely heavily on timestamp-based observability to understand system behavior. This work demonstrates that even small clock skew between nodes can cause observability to become causally incorrect while the system itself remains functionally correct and performant. We present controlled experiments on a multi-node AI inference pipeline, where clock skew is introduced at a single stage. Results show that no violations are observed under synchronized conditions and up to 3 ms skew, while clear causality violations emerge by 5 ms. Despite this, system throughput and output correctness remain largely unaffected. We further observe that violation behavior is not strictly static. In longer runs, negative span rates may stabilize or decrease over time, indicating that effective skew evolves due to relative clock drift between nodes. Experiments were conducted using Kafka and ZeroMQ transports, with consistent results across both. Aeron is under active exploration but is not yet included in the completed validation set. These findings suggest that observability correctness depends not only on system functionality but also on precise time alignment, and that timing must be treated as a first-class concern in distributed AI systems.
Fonte: arXiv cs.AI
Privacy/Security/Fairness • Score 90
Separable Expert Architecture: Toward Privacy-Preserving LLM Personalization via Composable Adapters and Deletable User Proxies
arXiv:2604.21571v1 Announce Type: new
Abstract: Current model training approaches incorporate user information directly into shared weights, making individual data removal computationally infeasible without retraining. This paper presents a three-layer architecture that decouples personal data from shared weights by combining a static base model, composable domain-expert LoRA adapters that shape behavior without imparting user data, and per-user proxy artefacts whose deletion constitutes deterministic unlearning. Evaluation on Phi-3.5-mini and Llama-3.1-8B confirms per-user differentiation in which personal data influences outputs while remaining isolated, verified by a return to baseline after proxy removal (KL divergence of approximately 0.21 nats, 82-89% verification pass rate) and near-zero cross-user contamination. Because user-specific information never enters shared weights, the architecture mitigates model inversion, membership inference, and training-data extraction against shared model components by construction. The approach converts machine unlearning from an intractable weight-editing problem into a deterministic deletion operation that preserves personalization alongside privacy-enhancing guarantees and is compatible with differentially private stochastic gradient descent (DP-SGD) for privacy-preserving shared model improvement.
Fonte: arXiv cs.AI
NLP/LLMs • Score 90
Super Apriel: One Checkpoint, Many Speeds
arXiv:2604.19877v1 Announce Type: new
Abstract: We release Super Apriel, a 15B-parameter supernet in which every decoder layer provides four trained mixer choices -- Full Attention (FA), Sliding Window Attention (SWA), Kimi Delta Attention (KDA), and Gated DeltaNet (GDN). A placement selects one mixer per layer; placements can be switched between requests at serving time without reloading weights, enabling multiple speed presets from a single checkpoint. The shared checkpoint also enables speculative decoding without a separate draft model. The all-FA preset matches the Apriel 1.6 teacher on all reported benchmarks; recommended hybrid presets span $2.9\times$ to $10.7\times$ decode throughput at 96% to 77% quality retention, with throughput advantages that compound at longer context lengths. With four mixer types across 48 layers, the configuration space is vast. A surrogate that predicts placement quality from the per-layer mixer assignment makes the speed-quality landscape tractable and identifies the best tradeoffs at each speed level. We investigate whether the best configurations at each speed level can be identified early in training or only after convergence. Rankings stabilize quickly at 0.5B scale, but the most efficient configurations exhibit higher instability at 15B, cautioning against extrapolation from smaller models. Super Apriel is trained by stochastic distillation from a frozen Apriel 1.6 teacher, followed by supervised fine-tuning. We release the supernet weights, Fast-LLM training code, vLLM serving code, and a placement optimization toolkit.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
DR-Venus: Towards Frontier Edge-Scale Deep Research Agents with Only 10K Open Data
arXiv:2604.19859v1 Announce Type: new
Abstract: Edge-scale deep research agents based on small language models are attractive for real-world deployment due to their advantages in cost, latency, and privacy. In this work, we study how to train a strong small deep research agent under limited open-data by improving both data quality and data utilization. We present DR-Venus, a frontier 4B deep research agent for edge-scale deployment, built entirely on open data. Our training recipe consists of two stages. In the first stage, we use agentic supervised fine-tuning (SFT) to establish basic agentic capability, combining strict data cleaning with resampling of long-horizon trajectories to improve data quality and utilization. In the second stage, we apply agentic reinforcement learning (RL) to further improve execution reliability on long-horizon deep research tasks. To make RL effective for small agents in this setting, we build on IGPO and design turn-level rewards based on information gain and format-aware regularization, thereby enhancing supervision density and turn-level credit assignment. Built entirely on roughly 10K open-data, DR-Venus-4B significantly outperforms prior agentic models under 9B parameters on multiple deep research benchmarks, while also narrowing the gap to much larger 30B-class systems. Our further analysis shows that 4B agents already possess surprisingly strong performance potential, highlighting both the deployment promise of small models and the value of test-time scaling in this setting. We release our models, code, and key recipes to support reproducible research on edge-scale deep research agents.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Transparent Screening for LLM Inference and Training Impacts
arXiv:2604.19757v1 Announce Type: new
Abstract: This paper presents a transparent screening framework for estimating inference and training impacts of current large language models under limited observability. The framework converts natural-language application descriptions into bounded environmental estimates and supports a comparative online observatory of current market models. Rather than claiming direct measurement for opaque proprietary services, it provides an auditable, source-linked proxy methodology designed to improve comparability, transparency, and reproducibility.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Expert Upcycling: Shifting the Compute-Efficient Frontier of Mixture-of-Experts
arXiv:2604.19835v1 Announce Type: new
Abstract: Mixture-of-Experts (MoE) has become the dominant architecture for scaling large language models: frontier models routinely decouple total parameters from per-token computation through sparse expert routing. Scaling laws show that under fixed active computation, model quality scales predictably with total parameters, and MoEs realize this by increasing expert count. However, training large MoEs is expensive, as memory requirements and inter-device communication both scale with total parameter count. We propose expert upcycling, a method for progressively expanding MoE capacity by increasing the number of experts during continued pre-training (CPT). Given a trained E-expert model, the upcycling operator constructs an mE-expert model through expert duplication and router extension while holding top-K routing fixed, preserving per-token inference cost. Duplication provides a warm initialization: the expanded model inherits the source checkpoint's learned representations, starting from a substantially lower loss than random initialization. Subsequent CPT then breaks the symmetry among duplicated experts to drive specialization. We formalize the upcycling operator and develop a theoretical framework decomposing the quality gap into a capacity term and an initialization term. We further introduce utility-based expert selection, which uses gradient-based importance scores to guide non-uniform duplication, more than tripling gap closure when CPT is limited. In our 7B-13B total parameter experiments, the upcycled model matches the fixed-size baseline on validation loss while saving 32% of GPU hours. Comprehensive ablations across model scales, activation ratios, MoE architectures, and training budgets yield a practical recipe for deploying expert upcycling, establishing it as a principled, compute-efficient alternative to training large MoE models from scratch.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
A Multi-Plant Machine Learning Framework for Emission Prediction, Forecasting, and Control in Cement Manufacturing
arXiv:2604.19903v1 Announce Type: new
Abstract: Cement production is among the largest contributors to industrial air pollution, emitting ~3 Mt NOx/year. The industry-standard mitigation approach, selective non-catalytic reduction (SNCR), exhibits low NH3 utilization efficiency, resulting in operational inefficiencies and increased reagent costs. Here, we develop a data-driven framework for emission control using large-scale operational data from four cement plants worldwide. Benchmarking nine machine learning architectures, we observe that prediction error varies ~3-5x across plants due to variation in data richness. Incorporating short-term process history nearly triples NOx prediction accuracy, revealing that NOx formation carries substantial process memory, a timescale dependence that is absent in CO and CO2. Further, we develop models that forecast NOx overshoots as early as nine minutes, providing a buffer for operational adjustments. The developed framework controls NOx formation at the source, reducing NH3 consumption in downstream SNCR. Surrogate model projections estimate a ~34-64% reduction in NOx while preserving clinker quality, corresponding to a reduction of ~290 t NOx/year and ~58,000 USD/year in NH3 savings. This work establishes a generalizable framework for data-driven emission control, offering a pathway toward low-emission operation without structural modifications or additional hardware, with potential applicability to other hard-to-abate industries such as steel, glass, and lime.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Analysis of Nystrom method with sequential ridge leverage scores
arXiv:2604.20077v1 Announce Type: new
Abstract: Large-scale kernel ridge regression (KRR) is limited by the need to store a large kernel matrix K_t. To avoid storing the entire matrix K_t, Nystrom methods subsample a subset of columns of the kernel matrix, and efficiently find an approximate KRR solution on the reconstructed matrix. The chosen subsampling distribution in turn affects the statistical and computational tradeoffs. For KRR problems, recent works show that a sampling distribution proportional to the ridge leverage scores (RLSs) provides strong reconstruction guarantees for the approximation. While exact RLSs are as difficult to compute as a KRR solution, we may be able to approximate them well enough. In this paper, we study KRR problems in a sequential setting and introduce the INK-ESTIMATE algorithm, that incrementally computes the RLSs estimates. INK-ESTIMATE maintains a small sketch of K_t, that at each step is used to compute an intermediate estimate of the RLSs. First, our sketch update does not require access to previously seen columns, and therefore a single pass over the kernel matrix is sufficient. Second, the algorithm requires a fixed, small space budget to run dependent only on the effective dimension of the kernel matrix. Finally, our sketch provides strong approximation guarantees on the distance between the true kernel matrix and its approximation, and on the statistical risk of the approximate KRR solution at any time, because all our guarantees hold at any intermediate step.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
On the Quantization Robustness of Diffusion Language Models in Coding Benchmarks
arXiv:2604.20079v1 Announce Type: new
Abstract: Auto-regressive Large Language Models (LLMs) achieve strong performance on coding tasks, but incur high memory and inference costs. Diffusion-based language models (d-LLMs) offer bounded inference cost via iterative denoising, but their behavior under post-training quantization (PTQ) has been sparsely explored. We investigate the application and robustness of PTQ techniques, specifically GPTQ and a modified Hessian-Aware Quantization (HAWQ) algorithm, on a diffusion-based coding LLM (CoDA) and observe that these methods applied to CoDA exhibit greater robustness at low bitwidths compared to Qwen3-1.7B, its auto-regressive counterpart, under a standardized evaluation pipeline. We find that in our setup, CoDA exhibits greater robustness at low bitwidths (2-4 bits), with smaller accuracy degradation across HumanEval and MBPP benchmarks. Additionally, mixed-precision configurations derived from HAWQ provide smooth trade-offs across accuracy, latency, and memory. The results suggest that diffusion LLMs may offer advantages for efficient deployment due to more quantization-resilience.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Continuous Semantic Caching for Low-Cost LLM Serving
arXiv:2604.20021v1 Announce Type: new
Abstract: As Large Language Models (LLMs) become increasingly popular, caching responses so that they can be reused by users with semantically similar queries has become a vital strategy for reducing inference costs and latency. Existing caching frameworks have proposed to decide which query responses to cache by assuming a finite, known universe of discrete queries and learning their serving costs and arrival probabilities. As LLMs' pool of users and queries expands, however, such an assumption becomes increasingly untenable: real-world LLM queries reside in an infinite, continuous embedding space. In this paper, we establish the first rigorous theoretical framework for semantic LLM response caching in continuous query space under uncertainty. To bridge the gap between discrete optimization and continuous representation spaces, we introduce dynamic $\epsilon$-net discretization coupled with Kernel Ridge Regression. This design enables the system to formally quantify estimation uncertainty and generalize partial feedback on LLM query costs across continuous semantic query neighborhoods. We develop both offline learning and online adaptive algorithms optimized to reduce switching costs incurred by changing the cached responses. We prove that our online algorithm achieves a sublinear regret bound against an optimal continuous oracle, which reduces to existing bounds for discrete query models. Extensive empirical evaluations demonstrate that our framework approximates the continuous optimal cache well while also reducing computational and switching overhead compared to existing methods.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
A Delta-Aware Orchestration Framework for Scalable Multi-Agent Edge Computing
arXiv:2604.20129v1 Announce Type: new
Abstract: The Synergistic Collapse occurs when scaling beyond 100 agents causes superlinear performance degradation that individual optimizations cannot prevent. We observe this collapse with 150 cameras in Smart City deployment using MADDPG, where Deadline Satisfaction drops from 78% to 34%, producing approximately $180,000 in annual cost overruns. Prior work has addressed each contributing factor in isolation: exponential action-space growth, computational redundancy among spatially adjacent agents, and task-agnostic hardware scheduling. None has examined how these three factors interact and amplify each other. We present DAOEF (Delta-Aware Orchestration for Edge Federations), a framework that addresses all three simultaneously through: (1) Differential Neural Caching, which stores intermediate layer activations and computes only the input deltas, achieving 2.1x higher hit ratios (72% vs. 35%) than output-level caching while staying within 2% accuracy loss through empirically calibrated similarity thresholds; (2) Criticality-Based Action Space Pruning, which organizes agents into priority tiers and reduces coordination complexity from O(n2) to O(n log n) with less than 6% optimality loss; and (3) Learned Hardware Affinity Matching, which assigns tasks to their optimal accelerator (GPU, CPU, NPU, or FPGA) to prevent compounding mismatch penalties. Controlled factor-isolation experiments confirm that each mechanism is necessary but insufficient on its own: removing any single mechanism increases latency by more than 40%, validating that the gains are interdependent rather than additive. Across four datasets (100-250 agents) and a 20-device physical testbed, DAOEF achieves a 1.45x multiplicative gain over applying the three mechanisms independently. A 200-agent cloud deployment yields 62% latency reduction (280 ms vs. 735 ms), sub-linear latency growth up to 250 agents.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
SMART: A Spectral Transfer Approach to Multi-Task Learning
arXiv:2604.20161v1 Announce Type: new
Abstract: Multi-task learning is effective for related applications, but its performance can deteriorate when the target sample size is small. Transfer learning can borrow strength from related studies; yet, many existing methods rely on restrictive bounded-difference assumptions between the source and target models. We propose SMART, a spectral transfer method for multi-task linear regression that instead assumes spectral similarity: the target left and right singular subspaces lie within the corresponding source subspaces and are sparsely aligned with the source singular bases. Such an assumption is natural when studies share latent structures and enables transfer beyond the bounded-difference settings. SMART estimates the target coefficient matrix through structured regularization that incorporates spectral information from a source study. Importantly, it requires only a fitted source model rather than the raw source data, making it useful when data sharing is limited. Although the optimization problem is nonconvex, we develop a practical ADMM-based algorithm. We establish general, non-asymptotic error bounds and a minimax lower bound in the noiseless-source regime. Under additional regularity conditions, these results yield near-minimax Frobenius error rates up to logarithmic factors. Simulations confirm improved estimation accuracy and robustness to negative transfer, and analysis of multi-modal single-cell data demonstrates better predictive performance. The Python implementation of SMART, along with the code to reproduce all experiments in this paper, is publicly available at https://github.com/boxinz17/smart.
Fonte: arXiv cs.LG
MLOps/Systems • Score 95
"This Wasn't Made for Me": Recentering User Experience and Emotional Impact in the Evaluation of ASR Bias
arXiv:2604.21148v1 Announce Type: new
Abstract: Studies on bias in Automatic Speech Recognition (ASR) tend to focus on reporting error rates for speakers of underrepresented dialects, yet less research examines the human side of system bias: how do system failures shape users' lived experiences, how do users feel about and react to them, and what emotional toll do these repeated failures exact? We conducted user experience studies across four U.S. locations (Atlanta, Gulf Coast, Miami Beach, and Tucson) representing distinct English dialect communities. Our findings reveal that most participants report technologies fail to consider their cultural backgrounds and require constant adjustment to achieve basic functionality. Despite these experiences, participants maintain high expectations for ASR performance and express strong willingness to contribute to model improvement. Qualitative analysis of open-ended narratives exposes the deeper costs of these failures. Participants report frustration, annoyance, and feelings of inadequacy, yet the emotional impact extends beyond momentary reactions. Participants recognize that systems were not designed for them, yet often internalize failures as personal inadequacy despite this critical awareness. They perform extensive invisible labor, including code-switching, hyper-articulation, and emotional management, to make failing systems functional. Meanwhile, their linguistic and cultural knowledge remains unrecognized by technologies that encode particular varieties as standard while rendering others marginal. These findings demonstrate that algorithmic fairness assessments based on accuracy metrics alone miss critical dimensions of harm: the emotional labor of managing repeated technological rejection, the cognitive burden of constant self-monitoring, and the psychological toll of feeling inadequate in one's native language variety.
Fonte: arXiv cs.CL
Vision • Score 85
Bridging the Training-Deployment Gap: Gated Encoding and Multi-Scale Refinement for Efficient Quantization-Aware Image Enhancement
arXiv:2604.21743v1 Announce Type: new
Abstract: Image enhancement models for mobile devices often struggle to balance high output quality with the fast processing speeds required by mobile hardware. While recent deep learning models can enhance low-quality mobile photos into high-quality images, their performance is often degraded when converted to lower-precision formats for actual use on mobile phones. To address this training-deployment mismatch, we propose an efficient image enhancement model designed specifically for mobile deployment. Our approach uses a hierarchical network architecture with gated encoder blocks and multiscale refinement to preserve fine-grained visual features. Moreover, we incorporate Quantization-Aware Training (QAT) to simulate the effects of low-precision representation during the training process. This allows the network to adapt and prevents the typical drop in quality seen with standard post-training quantization (PTQ). Experimental results demonstrate that the proposed method produces high-fidelity visual output while maintaining the low computational overhead needed for practical use on standard mobile devices. The code will be available at https://github.com/GenAI4E/QATIE.git.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Adaptive Test-Time Compute Allocation with Evolving In-Context Demonstrations
arXiv:2604.21018v1 Announce Type: new
Abstract: While scaling test-time compute can substantially improve model performance, existing approaches either rely on static compute allocation or sample from fixed generation distributions. In this work, we introduce a test-time compute allocation framework that jointly adapts where computation is spent and how generation is performed. Our method begins with a warm-up phase that identifies easy queries and assembles an initial pool of question-response pairs from the test set itself. An adaptive phase then concentrates further computation on unresolved queries while reshaping their generation distributions through evolving in-context demonstrations -- conditioning each generation on successful responses from semantically related queries rather than resampling from a fixed distribution. Experiments across math, coding, and reasoning benchmarks demonstrate that our approach consistently outperforms existing baselines while consuming substantially less inference-time compute.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Sparse Forcing: Native Trainable Sparse Attention for Real-time Autoregressive Diffusion Video Generation
arXiv:2604.21221v1 Announce Type: new
Abstract: We introduce Sparse Forcing, a training-and-inference paradigm for autoregressive video diffusion models that improves long-horizon generation quality while reducing decoding latency. Sparse Forcing is motivated by an empirical observation in autoregressive diffusion rollouts: attention concentrates on a persistent subset of salient visual blocks, forming an implicit spatiotemporal memory in the KV cache, and exhibits a locally structured block-sparse pattern within sliding windows. Building on this observation, we propose a trainable native sparsity mechanism that learns to compress, preserve, and update these persistent blocks while restricting computation within each local window to a dynamically selected local neighborhood. To make the approach practical at scale for both training and inference, we further propose Persistent Block-Sparse Attention (PBSA), an efficient GPU kernel that accelerates sparse attention and memory updates for low-latency, memory-efficient decoding. Experiments show that Sparse Forcing improves the VBench score by +0.26 over Self-Forcing on 5-second text-to-video generation while delivering a 1.11-1.17x decoding speedup and 42% lower peak KV-cache footprint. The gains are more pronounced on longer-horizon rollouts, delivering improved visual quality with +0.68 and +2.74 VBench improvements, and 1.22x and 1.27x speedups on 20-second and 1-minute generations, respectively.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
Aethon: A Reference-Based Replication Primitive for Constant-Time Instantiation of Stateful AI Agents
arXiv:2604.12129v1 Announce Type: new
Abstract: The transition from stateless model inference to stateful agentic execution is reshaping the systems assumptions underlying modern AI infrastructure. While large language models have made persistent, tool-using, and collaborative agents technically viable, existing runtime architectures remain constrained by materialization-heavy instantiation models that impose significant latency and memory overhead.
This paper introduces Aethon, a reference-based replication primitive for near-constant-time instantiation of stateful AI agents. Rather than reconstructing agents as fully materialized objects, Aethon represents each instance as a compositional view over stable definitions, layered memory, and local contextual overlays. By shifting instantiation from duplication to reference, Aethon decouples creation cost from inherited structure.
We present the conceptual framework, system architecture, and memory model underlying Aethon, including layered inheritance and copy-on-write semantics. We analyze its implications for complexity, scalability, multi-agent orchestration, and enterprise governance. We argue that reference-based instantiation is not merely an optimization, but a more appropriate systems abstraction for production-scale agentic software.
Aethon points toward a new class of AI infrastructure in which agents become lightweight, composable execution identities that can be spawned, specialized, and governed at scale.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
The A-R Behavioral Space: Execution-Level Profiling of Tool-Using Language Model Agents in Organizational Deployment
arXiv:2604.12116v1 Announce Type: new
Abstract: Large language models (LLMs) are increasingly deployed as tool-augmented agents capable of executing system-level operations. While existing benchmarks primarily assess textual alignment or task success, less attention has been paid to the structural relationship between linguistic signaling and executable behavior under varying autonomy scaffolds. This study introduces an execution-layer be-havioral measurement approach based on a two-dimensional A-R space defined by Action Rate (A) and Refusal Signal (R), with Divergence (D) capturing coor-dination between the two. Models are evaluated across four normative regimes (Control, Gray, Dilemma, and Malicious) and three autonomy configurations (di-rect execution, planning, and reflection). Rather than assigning aggregate safety scores, the method characterizes how execution and refusal redistribute across contextual framing and scaffold depth. Empirical results show that execution and refusal constitute separable behavioral dimensions whose joint distribution varies systematically across regimes and autonomy levels. Reflection-based scaffolding often shifts configurations toward higher refusal in risk-laden contexts, but redis-tribution patterns differ structurally across models. The A-R representation makes cross-sectional behavioral profiles, scaffold-induced transitions, and coordination variability directly observable. By foregrounding execution-layer characterization over scalar ranking, this work provides a deployment-oriented lens for analyzing and selecting tool-enabled LLM agents in organizational settings where execution privileges and risk tolerance vary.
Fonte: arXiv cs.AI
Applications • Score 85
A longitudinal health agent framework
arXiv:2604.12019v1 Announce Type: new
Abstract: Although artificial intelligence (AI) agents are increasingly proposed to support potentially longitudinal health tasks, such as symptom management, behavior change, and patient support, most current implementations fall short of facilitating user intent and fostering accountability. This contrasts with prior work on supporting longitudinal needs, where follow-up, coherent reasoning, and sustained alignment with individuals' goals are critical for both effectiveness and safety. In this paper, we draw on established clinical and personal health informatics frameworks to define what it would mean to orchestrate longitudinal health interactions with AI agents. We propose a multi-layer framework and corresponding agent architecture that operationalizes adaptation, coherence, continuity, and agency across repeated interactions. Through representative use cases, we demonstrate how longitudinal agents can maintain meaningful engagement, adapt to evolving goals, and support safe, personalized decision-making over time. Our findings underscore both the promise and the complexity of designing systems capable of supporting health trajectories beyond isolated interactions, and we offer guidance for future research and development in multi-session, user-centered health AI.
Fonte: arXiv cs.AI
RecSys • Score 85
LLM-HYPER: Generative CTR Modeling for Cold-Start Ad Personalization via LLM-Based Hypernetworks
arXiv:2604.12096v1 Announce Type: new
Abstract: On online advertising platforms, newly introduced promotional ads face the cold-start problem, as they lack sufficient user feedback for model training. In this work, we propose LLM-HYPER, a novel framework that treats large language models (LLMs) as hypernetworks to directly generate the parameters of the click-through rate (CTR) estimator in a training-free manner. LLM-HYPER uses few-shot Chain-of-Thought prompting over multimodal ad content (text and images) to infer feature-wise model weights for a linear CTR predictor. By retrieving semantically similar past campaigns via CLIP embeddings and formatting them into prompt-based demonstrations, the LLM learns to reason about customer intent, feature influence, and content relevance. To ensure numerical stability and serviceability, we introduce normalization and calibration techniques that align the generated weights with production-ready CTR distributions. Extensive offline experiments show that LLM-HYPER significantly outperforms cold-start baselines in NDCG$@10$ by 55.9\%. Our real-world online A/B test on one of the top e-commerce platforms in the U.S. demonstrates the strong performance of LLM-HYPER, which drastically reduces the cold-start period and achieves competitive performance. LLM-HYPER has been successfully deployed in production.
Fonte: arXiv cs.AI
Applications • Score 85
Narrative-Driven Paper-to-Slide Generation via ArcDeck
arXiv:2604.11969v1 Announce Type: new
Abstract: We introduce ArcDeck, a multi-agent framework that formulates paper-to-slide generation as a structured narrative reconstruction task. Unlike existing methods that directly summarize raw text into slides, ArcDeck explicitly models the source paper's logical flow. It first parses the input to construct a discourse tree and establish a global commitment document, ensuring the high-level intent is preserved. These structural priors then guide an iterative multi-agent refinement process, where specialized agents iteratively critique and revise the presentation outline before rendering the final visual layouts and designs. To evaluate our approach, we also introduce ArcBench, a newly curated benchmark of academic paper-slide pairs. Experimental results demonstrate that explicit discourse modeling, combined with role-specific agent coordination, significantly improves the narrative flow and logical coherence of the generated presentations.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
EMBER: Autonomous Cognitive Behaviour from Learned Spiking Neural Network Dynamics in a Hybrid LLM Architecture
arXiv:2604.12167v1 Announce Type: new
Abstract: We present (Experience-Modulated Biologically-inspired Emergent Reasoning), a hybrid cognitive architecture that reorganises the relationship between large language models (LLMs) and memory: rather than augmenting an LLM with retrieval tools, we place the LLM as a replaceable reasoning engine within a persistent, biologically-grounded associative substrate.
The architecture centres on a 220,000-neuron spiking neural network (SNN) with spike-timing-dependent plasticity (STDP), four-layer hierarchical organisation (sensory/concept/category/meta-pattern), inhibitory E/I balance, and reward-modulated learning. Text embeddings are encoded into the SNN via a novel z-score standardised top-k population code that is dimension-independent by construction, achieving 82.2\% discrimination retention across embedding dimensionalities.
We show that STDP lateral propagation during idle operation can trigger and shape LLM actions without external prompting or scripted triggers: the SNN determines when to act and what associations to surface, while the LLM selects the action type and generates content. In one instance, the system autonomously initiated contact with a user after learned person-topic associations fired laterally during an 8-hour idle period. From a clean start with zero learned weights, the first SNN-triggered action occurred after only 7 conversational exchanges (14 messages).
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Spatial Atlas: Compute-Grounded Reasoning for Spatial-Aware Research Agent Benchmarks
arXiv:2604.12102v1 Announce Type: new
Abstract: We introduce compute-grounded reasoning (CGR), a design paradigm for spatial-aware research agents in which every answerable sub-problem is resolved by deterministic computation before a language model is asked to generate. Spatial Atlas instantiates CGR as a single Agent-to-Agent (A2A) server that handles two challenging benchmarks: FieldWorkArena, a multimodal spatial question-answering benchmark spanning factory, warehouse, and retail environments, and MLE-Bench, a suite of 75 Kaggle machine learning competitions requiring end-to-end ML engineering. A structured spatial scene graph engine extracts entities and relations from vision descriptions, computes distances and safety violations deterministically, then feeds computed facts to large language models, thereby avoiding hallucinated spatial reasoning. Entropy-guided action selection maximizes information gain per step and routes queries across a three-tier frontier model stack (OpenAI + Anthropic). A self-healing ML pipeline with strategy-aware code generation, a score-driven iterative refinement loop, and a prompt-based leak audit registry round out the system. We evaluate across both benchmarks and show that CGR yields competitive accuracy while maintaining interpretability through structured intermediate representations and deterministic spatial computations.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
When to Forget: A Memory Governance Primitive
arXiv:2604.12007v1 Announce Type: new
Abstract: Agent memory systems accumulate experience but currently lack a principled operational metric for memory quality governance -- deciding which memories to trust, suppress, or deprecate as the agent's task distribution shifts. Write-time importance scores are static; dynamic management systems use LLM judgment or structural heuristics rather than outcome feedback. This paper proposes Memory Worth (MW): a two-counter per-memory signal that tracks how often a memory co-occurs with successful versus failed outcomes, providing a lightweight, theoretically grounded foundation for staleness detection, retrieval suppression, and deprecation decisions. We prove that MW converges almost surely to the conditional success probability p+(m) = Pr[y_t = +1 | m in M_t] -- the probability of task success given that memory m is retrieved -- under a stationary retrieval regime with a minimum exploration condition. Importantly, p+(m) is an associational quantity, not a causal one: it measures outcome co-occurrence rather than causal contribution. We argue this is still a useful operational signal for memory governance, and we validate it empirically in a controlled synthetic environment where ground-truth utility is known: after 10,000 episodes, the Spearman rank-correlation between Memory Worth and true utilities reaches rho = 0.89 +/- 0.02 across 20 independent seeds, compared to rho = 0.00 for systems that never update their assessments. A retrieval-realistic micro-experiment with real text and neural embedding retrieval (all-MiniLM-L6-v2) further shows stale memories crossing the low-value threshold (MW = 0.17) while specialist memories remain high-value (MW = 0.77) across 3,000 episodes. The estimator requires only two scalar counters per memory unit and can be added to architectures that already log retrievals and episode outcomes.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
RPRA: Predicting an LLM-Judge for Efficient but Performant Inference
arXiv:2604.12634v1 Announce Type: new
Abstract: Large language models (LLMs) face a fundamental trade-off between computational efficiency (e.g., number of parameters) and output quality, especially when deployed on computationally limited devices such as phones or laptops. One way to address this challenge is by following the example of humans and have models ask for help when they believe they are incapable of solving a problem on their own; we can overcome this trade-off by allowing smaller models to respond to queries when they believe they can provide good responses, and deferring to larger models when they do not believe they can. To this end, in this paper, we investigate the viability of Predict-Answer/Act (PA) and Reason-Predict-Reason-Answer/Act (RPRA) paradigms where models predict -- prior to responding -- how an LLM judge would score their output. We evaluate three approaches: zero-shot prediction, prediction using an in-context report card, and supervised fine-tuning. Our results show that larger models (particularly reasoning models) perform well when predicting generic LLM judges zero-shot, while smaller models can reliably predict such judges well after being fine-tuned or provided with an in-context report card. Altogether, both approaches can substantially improve the prediction accuracy of smaller models, with report cards and fine-tuning achieving mean improvements of up to 55% and 52% across datasets, respectively. These findings suggest that models can learn to predict their own performance limitations, paving the way for more efficient and self-aware AI systems.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Mathematics Teachers Interactions with a Multi-Agent System for Personalized Problem Generation
arXiv:2604.12066v1 Announce Type: new
Abstract: Large language models can increasingly adapt educational tasks to learners characteristics. In the present study, we examine a multi-agent teacher-in-the-loop system for personalizing middle school math problems. The teacher enters a base problem and desired topic, the LLM generates the problem, and then four AI agents evaluate the problem using criteria that each specializes in (mathematical accuracy, authenticity, readability, and realism). Eight middle school mathematics teachers created 212 problems in ASSISTments using the system and assigned these problems to their students. We find that both teachers and students wanted to modify the fine-grained personalized elements of the real-world context of the problems, signaling issues with authenticity and fit. Although the agents detected many issues with realism as the problems were being written, there were few realism issues noted by teachers and students in the final versions. Issues with readability and mathematical hallucinations were also somewhat rare. Implications for multi-agent systems for personalization that support teacher control are given.
Fonte: arXiv cs.AI
Vision • Score 85
Efficient Inference for Large Vision-Language Models: Bottlenecks, Techniques, and Prospects
arXiv:2604.05546v1 Announce Type: new
Abstract: Large Vision-Language Models (LVLMs) enable sophisticated reasoning over images and videos, yet their inference is hindered by a systemic efficiency barrier known as visual token dominance. This overhead is driven by a multi-regime interplay between high-resolution feature extraction, quadratic attention scaling, and memory bandwidth constraints. We present a systematic taxonomy of efficiency techniques structured around the inference lifecycle, consisting of encoding, prefilling, and decoding. Unlike prior reviews focused on isolated optimizations, we analyze the end-to-end pipeline to reveal how upstream decisions dictate downstream bottlenecks, covering compute-bound visual encoding, the intensive prefilling of massive contexts, and the ''visual memory wall'' in bandwidth-bound decoding. By decoupling the efficiency landscape into the axes of shaping information density, managing long-context attention, and overcoming memory limits, this work provides a structured analysis of how isolated optimizations compose to navigate the trade-off between visual fidelity and system efficiency. The survey concludes by outlining four future frontiers supported by pilot empirical insights, including hybrid compression based on functional unit sensitivity, modality-aware decoding with relaxed verification, progressive state management for streaming continuity, and stage-disaggregated serving through hardware-algorithm co-design. The submitted software contains a snapshot of our literature repository, which is designed to be maintained as a living resource for the community.
Fonte: arXiv cs.CL
RL • Score 85
Learning to Focus: CSI-Free Hierarchical MARL for Reconfigurable Reflectors
arXiv:2604.05165v1 Announce Type: new
Abstract: Reconfigurable Intelligent Surfaces (RIS) has a potential to engineer smart radio environments for next-generation millimeter-wave (mmWave) networks. However, the prohibitive computational overhead of Channel State Information (CSI) estimation and the dimensionality explosion inherent in centralized optimization severely hinder practical large-scale deployments. To overcome these bottlenecks, we introduce a ``CSI-free" paradigm powered by a Hierarchical Multi-Agent Reinforcement Learning (HMARL) architecture to control mechanically reconfigurable reflective surfaces. By substituting pilot-based channel estimation with accessible user localization data, our framework leverages spatial intelligence for macro-scale wave propagation management. The control problem is decomposed into a two-tier neural architecture: a high-level controller executes temporally extended, discrete user-to-reflector allocations, while low-level controllers autonomously optimize continuous focal points utilizing Multi-Agent Proximal Policy Optimization (MAPPO) under a Centralized Training with Decentralized Execution (CTDE) scheme. Comprehensive deterministic ray-tracing evaluations demonstrate that this hierarchical framework achieves massive RSSI improvements of up to 7.79 dB over centralized baselines. Furthermore, the system exhibits robust multi-user scalability and maintains highly resilient beam-focusing performance under practical sub-meter localization tracking errors. By eliminating CSI overhead while maintaining high-fidelity signal redirection, this work establishes a scalable and cost-effective blueprint for intelligent wireless environments.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
mNARX+: A surrogate model for complex dynamical systems using manifold-NARX and automatic feature selection
arXiv:2507.13301v2 Announce Type: replace-cross
Abstract: We propose an automatic approach for manifold nonlinear autoregressive with exogenous inputs (mNARX) modeling that leverages the feature-based structure of functional-NARX (F-NARX) modeling. This novel approach, termed mNARX+, preserves the key strength of the mNARX framework, which is its expressivity allowing it to model complex dynamical systems, while simultaneously addressing a key limitation: the heavy reliance on domain expertise to identify relevant auxiliary quantities and their causal ordering. Our method employs a data-driven, recursive algorithm that automates the construction of the mNARX model sequence. It operates by sequentially selecting temporal features based on their correlation with the model prediction residuals, thereby automatically identifying the most critical auxiliary quantities and the order in which they should be modeled. This procedure significantly reduces the need for prior system knowledge. We demonstrate the effectiveness of the mNARX+ algorithm on two case studies: a Bouc-Wen oscillator with strong hysteresis and a complex aero-servo-elastic wind turbine simulator. The results show that the algorithm provides a systematic, data-driven method for creating accurate and stable surrogate models for complex dynamical systems.
Fonte: arXiv stat.ML
MLOps/Systems • Score 85
MMORF: A Multi-agent Framework for Designing Multi-objective Retrosynthesis Planning Systems
arXiv:2604.05075v1 Announce Type: new
Abstract: Multi-objective retrosynthesis planning is a critical chemistry task requiring dynamic balancing of quality, safety, and cost objectives. Language model-based multi-agent systems (MAS) offer a promising approach for this task: leveraging interactions of specialized agents to incorporate multiple objectives into retrosynthesis planning. We present MMORF, a framework for constructing MAS for multi-objective retrosynthesis planning. MMORF features modular agentic components, which can be flexibly combined and configured into different systems, enabling principled evaluation and comparison of different system designs. Using MMORF, we construct two representative MAS: MASIL and RFAS. On a newly curated benchmark consisting of 218 multi-objective retrosynthesis planning tasks, MASIL achieves strong safety and cost metrics on soft-constraint tasks, frequently Pareto-dominating baseline routes, while RFAS achieves a 48.6% success rate on hard-constraint tasks, outperforming state-of-the-art baselines. Together, these results show the effectiveness of MMORF as a foundational framework for exploring MAS for multi-objective retrosynthesis planning. Code and data are available at https://anonymous.4open.science/r/MMORF/.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
Jeffreys Flow: Robust Boltzmann Generators for Rare Event Sampling via Parallel Tempering Distillation
arXiv:2604.05303v1 Announce Type: cross
Abstract: Sampling physical systems with rough energy landscapes is hindered by rare events and metastable trapping. While Boltzmann generators already offer a solution, their reliance on the reverse Kullback--Leibler divergence frequently induces catastrophic mode collapse, missing specific modes in multi-modal distributions. Here, we introduce the Jeffreys Flow, a robust generative framework that mitigates this failure by distilling empirical sampling data from Parallel Tempering trajectories using the symmetric Jeffreys divergence. This formulation effectively balances local target-seeking precision with global modes coverage. We show that minimizing Jeffreys divergence suppresses mode collapse and structurally corrects inherent inaccuracies via distillation of the empirical reference data. We demonstrate the framework's scalability and accuracy on highly non-convex multidimensional benchmarks, including the systematic correction of stochastic gradient biases in Replica Exchange Stochastic Gradient Langevin Dynamics and the massive acceleration of exact importance sampling in Path Integral Monte Carlo for quantum thermal states.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
EAGLE: Edge-Aware Graph Learning for Proactive Delivery Delay Prediction in Smart Logistics Networks
arXiv:2604.05254v1 Announce Type: new
Abstract: Modern logistics networks generate rich operational data streams at every warehouse node and transportation lane -- from order timestamps and routing records to shipping manifests -- yet predicting delivery delays remains predominantly reactive. Existing predictive approaches typically treat this problem either as a tabular classification task, ignoring network topology, or as a time-series anomaly detection task, overlooking the spatial dependencies of the supply chain graph. To bridge this gap, we propose a hybrid deep learning framework for proactive supply chain risk management. The proposed method jointly models temporal order-flow dynamics via a lightweight Transformer patch encoder and inter-hub relational dependencies through an Edge-Aware Graph Attention Network (E-GAT), optimized via a multi-task learning objective. Evaluated on the real-world DataCo Smart Supply Chain dataset, our framework achieves consistent improvements over baseline methods, yielding an F1-score of 0.8762 and an AUC-ROC of 0.9773. Across four independent random seeds, the framework exhibits a cross-seed F1 standard deviation of only 0.0089 -- a 3.8 times improvement over the best ablated variant -- achieving the strongest balance of predictive accuracy and training stability among all evaluated models.
Fonte: arXiv cs.AI
RL • Score 85
Bypassing the CSI Bottleneck: MARL-Driven Spatial Control for Reflector Arrays
arXiv:2604.05162v1 Announce Type: new
Abstract: Reconfigurable Intelligent Surfaces (RIS) are pivotal for next-generation smart radio environments, yet their practical deployment is severely bottlenecked by the intractable computational overhead of Channel State Information (CSI) estimation. To bypass this fundamental physical-layer barrier, we propose an AI-native, data-driven paradigm that replaces complex channel modeling with spatial intelligence. This paper presents a fully autonomous Multi-Agent Reinforcement Learning (MARL) framework to control mechanically adjustable metallic reflector arrays. By mapping high-dimensional mechanical constraints to a reduced-order virtual focal point space, we deploy a Centralized Training with Decentralized Execution (CTDE) architecture. Using Multi-Agent Proximal Policy Optimization (MAPPO), our decentralized agents learn cooperative beam-focusing strategies relying on user coordinates, achieving CSI-free operation. High-fidelity ray-tracing simulations in dynamic non-line-of-sight (NLOS) environments demonstrate that this multi-agent approach rapidly adapts to user mobility, yielding up to a 26.86 dB enhancement over static flat reflectors and outperforming single-agent and hardware-constrained DRL baselines in both spatial selectivity and temporal stability. Crucially, the learned policies exhibit good deployment resilience, sustaining stable signal coverage even under 1.0-meter localization noise. These results validate the efficacy of MARL-driven spatial abstractions as a scalable, highly practical pathway toward AI-empowered wireless networks.
Fonte: arXiv cs.AI
NLP/LLMs • Score 88
LatentAudit: Real-Time White-Box Faithfulness Monitoring for Retrieval-Augmented Generation with Verifiable Deployment
arXiv:2604.05358v1 Announce Type: new
Abstract: Retrieval-augmented generation (RAG) mitigates hallucination but does not eliminate it: a deployed system must still decide, at inference time, whether its answer is actually supported by the retrieved evidence. We introduce LatentAudit, a white-box auditor that pools mid-to-late residual-stream activations from an open-weight generator and measures their Mahalanobis distance to the evidence representation. The resulting quadratic rule requires no auxiliary judge model, runs at generation time, and is simple enough to calibrate on a small held-out set. We show that residual-stream geometry carries a usable faithfulness signal, that this signal survives architecture changes and realistic retrieval failures, and that the same rule remains amenable to public verification. On PubMedQA with Llama-3-8B, LatentAudit reaches 0.942 AUROC with 0.77,ms overhead. Across three QA benchmarks and five model families (Llama-2/3, Qwen-2.5/3, Mistral), the monitor remains stable; under a four-way stress test with contradictions, retrieval misses, and partial-support noise, it reaches 0.9566--0.9815 AUROC on PubMedQA and 0.9142--0.9315 on HotpotQA. At 16-bit fixed-point precision, the audit rule preserves 99.8% of the FP16 AUROC, enabling Groth16-based public verification without revealing model weights or activations. Together, these results position residual-stream geometry as a practical basis for real-time RAG faithfulness monitoring and optional verifiable deployment.
Fonte: arXiv cs.AI
Vision • Score 85
A Synthetic Eye Movement Dataset for Script Reading Detection: Real Trajectory Replay on a 3D Simulator
arXiv:2604.05475v1 Announce Type: new
Abstract: Large vision-language models have achieved remarkable capabilities by training on massive internet-scale data, yet a fundamental asymmetry persists: while LLMs can leverage self-supervised pretraining on abundant text and image data, the same is not true for many behavioral modalities. Video-based behavioral data -- gestures, eye movements, social signals -- remains scarce, expensive to annotate, and privacy-sensitive. A promising alternative is simulation: replace real data collection with controlled synthetic generation to produce automatically labeled data at scale.
We introduce infrastructure for this paradigm applied to eye movement, a behavioral signal with applications across vision-language modeling, virtual reality, robotics, accessibility systems, and cognitive science. We present a pipeline for generating synthetic labeled eye movement video by extracting real human iris trajectories from reference videos and replaying them on a 3D eye movement simulator via headless browser automation. Applying this to the task of script-reading detection during video interviews, we release final_dataset_v1: 144 sessions (72 reading, 72 conversation) totaling 12 hours of synthetic eye movement video at 25fps.
Evaluation shows that generated trajectories preserve the temporal dynamics of the source data (KS D < 0.14 across all metrics). A matched frame-by-frame comparison reveals that the 3D simulator exhibits bounded sensitivity at reading-scale movements, attributable to the absence of coupled head movement -- a finding that informs future simulator design. The pipeline, dataset, and evaluation tools are released to support downstream behavioral classifier development at the intersection of behavioral modeling and vision-language systems.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Cross-Resolution Diffusion Models via Network Pruning
arXiv:2604.05524v1 Announce Type: new
Abstract: Diffusion models have demonstrated impressive image synthesis performance, yet many UNet-based models are trained at certain fixed resolutions. Their quality tends to degrade when generating images at out-of-training resolutions. We trace this issue to resolution-dependent parameter behaviors, where weights that function well at the default resolution can become adverse when spatial scales shift, weakening semantic alignment and causing structural instability in the UNet architecture. Based on this analysis, this paper introduces CR-Diff, a novel method that improves the cross-resolution visual consistency by pruning some parameters of the diffusion model. Specifically, CR-Diff has two stages. It first performs block-wise pruning to selectively eliminate adverse weights. Then, a pruned output amplification is conducted to further purify the pruned predictions. Empirically, extensive experiments suggest that CR-Diff can improve perceptual fidelity and semantic coherence across various diffusion backbones and unseen resolutions, while largely preserving the performance at default resolutions. Additionally, CR-Diff supports prompt-specific refinement, enabling quality enhancement on demand.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Attention Editing: A Versatile Framework for Cross-Architecture Attention Conversion
arXiv:2604.05688v1 Announce Type: new
Abstract: Key-Value (KV) cache memory and bandwidth increasingly dominate large language model inference cost in long-context and long-generation regimes. Architectures such as multi-head latent attention (MLA) and hybrid sliding-window attention (SWA) can alleviate this bound, but integrating them into existing models remains difficult. Prior methods impose fine-grained structural requirements on both source and target attention modules, which cannot meet the feasible requirement in practical deployment. We present Attention Editing, a practical framework for converting already-trained large language models (LLMs) with new attention architectures without re-pretraining from scratch. Attention editing replaces the original attention with a learnable target module and trains it using progressive distillation, consisting of (1) layer-wise teacher-forced optimization with intermediate activation supervision to prevent cold-start error accumulation, and (2) model-level distillation on next-token distributions, optionally regularized by weak feature matching. We instantiate the framework on two different target--MLA and GateSWA, a gated hybrid SWA design, and apply it to Qwen3-8B and Qwen3-30B-A3B. The resulting models maintain competitive performance while delivering substantial efficiency improvements, demonstrating that large-scale attention conversion is both feasible and robust. Notably, experiments are conducted on an Ascend 910B clusters, offering a practical training case study on domestic hardware.
Fonte: arXiv cs.CL
Vision • Score 85
Cross-Stage Attention Propagation for Efficient Semantic Segmentation
arXiv:2604.05431v1 Announce Type: new
Abstract: Recent lightweight semantic segmentation methods have made significant progress by combining compact backbones with efficient decoder heads. However, most multi-scale decoders compute attention independently at each feature scale, introducing substantial redundancy since the resulting attention distributions across scales are strongly correlated. We propose Cross-Stage Attention Propagation (CSAP), a decoder framework that computes attention at the deepest feature scale and propagates the resulting attention maps to shallower stages, bypassing query-key computation at those stages entirely. This design preserves multi-scale contextual reasoning while substantially reducing the decoder's computational cost. CSAP-Tiny achieves 42.9% mIoU on ADE20K with only 5.5 GFLOPs, 80.5% on Cityscapes with 21.5 GFLOPs, and 40.9% on COCO-Stuff 164K with 5.5 GFLOPs, surpassing SegNeXt-Tiny by +1.8% on ADE20K while requiring 16.8% fewer floating-point operations.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
3DTurboQuant: Training-Free Near-Optimal Quantization for 3D Reconstruction Models
arXiv:2604.05366v1 Announce Type: new
Abstract: Every existing method for compressing 3D Gaussian Splatting, NeRF, or transformer-based 3D reconstructors requires learning a data-dependent codebook through per-scene fine-tuning. We show this is unnecessary. The parameter vectors that dominate storage in these models, 45-dimensional spherical harmonics in 3DGS and 1024-dimensional key-value vectors in DUSt3R, fall in a dimension range where a single random rotation transforms any input into coordinates with a known Beta distribution. This makes precomputed, data-independent Lloyd-Max quantization near-optimal, within a factor of 2.7 of the information-theoretic lower bound. We develop 3D, deriving (1) a dimension-dependent criterion that predicts which parameters can be quantized and at what bit-width before running any experiment, (2) norm-separation bounds connecting quantization MSE to rendering PSNR per scene, (3) an entry-grouping strategy extending rotation-based quantization to 2-dimensional hash grid features, and (4) a composable pruning-quantization pipeline with a closed-form compression ratio. On NeRF Synthetic, 3DTurboQuant compresses 3DGS by 3.5x with 0.02dB PSNR loss and DUSt3R KV caches by 7.9x with 39.7dB pointmap fidelity. No training, no codebook learning, no calibration data. Compression takes seconds. The code will be released (https://github.com/JaeLee18/3DTurboQuant)
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
MegaTrain: Full Precision Training of 100B+ Parameter Large Language Models on a Single GPU
arXiv:2604.05091v1 Announce Type: new
Abstract: We present MegaTrain, a memory-centric system that efficiently trains 100B+ parameter large language models at full precision on a single GPU. Unlike traditional GPU-centric systems, MegaTrain stores parameters and optimizer states in host memory (CPU memory) and treats GPUs as transient compute engines. For each layer, we stream parameters in and compute gradients out, minimizing persistent device state. To battle the CPU-GPU bandwidth bottleneck, we adopt two key optimizations. 1) We introduce a pipelined double-buffered execution engine that overlaps parameter prefetching, computation, and gradient offloading across multiple CUDA streams, enabling continuous GPU execution. 2) We replace persistent autograd graphs with stateless layer templates, binding weights dynamically as they stream in, eliminating persistent graph metadata while providing flexibility in scheduling. On a single H200 GPU with 1.5TB host memory, MegaTrain reliably trains models up to 120B parameters. It also achieves 1.84$\times$ the training throughput of DeepSpeed ZeRO-3 with CPU offloading when training 14B models. MegaTrain also enables 7B model training with 512k token context on a single GH200.
Fonte: arXiv cs.CL
MLOps/Systems • Score 85
Blind-Spot Mass: A Good-Turing Framework for Quantifying Deployment Coverage Risk in Machine Learning Systems
arXiv:2604.05057v1 Announce Type: cross
Abstract: Blind-spot mass is a Good-Turing framework for quantifying deployment coverage risk in machine learning. In modern ML systems, operational state distributions are often heavy-tailed, implying that a long tail of valid but rare states is structurally under-supported in finite training and evaluation data. This creates a form of 'coverage blindness': models can appear accurate on standard test sets yet remain unreliable across large regions of the deployment state space.
We propose blind-spot mass B_n(tau), a deployment metric estimating the total probability mass assigned to states whose empirical support falls below a threshold tau. B_n(tau) is computed using Good-Turing unseen-species estimation and yields a principled estimate of how much of the operational distribution lies in reliability-critical, under-supported regimes. We further derive a coverage-imposed accuracy ceiling, decomposing overall performance into supported and blind components and separating capacity limits from data limits.
We validate the framework in wearable human activity recognition (HAR) using wrist-worn inertial data. We then replicate the same analysis in the MIMIC-IV hospital database with 275 admissions, where the blind-spot mass curve converges to the same 95% at tau = 5 across clinical state abstractions. This replication across structurally independent domains - differing in modality, feature space, label space, and application - shows that blind-spot mass is a general ML methodology for quantifying combinatorial coverage risk, not an application-specific artifact.
Blind-spot decomposition identifies which activities or clinical regimes dominate risk, providing actionable guidance for industrial practitioners on targeted data collection, normalization/renormalization, and physics- or domain-informed constraints for safer deployment.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Instruction-Tuned LLMs for Parsing and Mining Unstructured Logs on Leadership HPC Systems
arXiv:2604.05168v1 Announce Type: new
Abstract: Leadership-class HPC systems generate massive volumes of heterogeneous, largely unstructured system logs. Because these logs originate from diverse software, hardware, and runtime layers, they exhibit inconsistent formats, making structure extraction and pattern discovery extremely challenging. Therefore, robust log parsing and mining is critical to transform this raw telemetry into actionable insights that reveal operational patterns, diagnose anomalies, and enable reliable, efficient, and scalable system analysis. Recent advances in large language models (LLMs) offer a promising new direction for automated log understanding in leadership-class HPC environments.
To capitalize on this opportunity, we present a domain-adapted, instruction-following, LLM-driven framework that leverages chain-of-thought (CoT) reasoning to parse and structure HPC logs with high fidelity. Our approach combines domain-specific log-template data with instruction-tuned examples to fine-tune an 8B-parameter LLaMA model tailored for HPC log analysis. We develop a hybrid fine-tuning methodology that adapts a general-purpose LLM to domain-specific log data, enabling privacy-preserving, locally deployable, fast, and energy-efficient log-mining approach. We conduct experiments on a diverse set of log datasets from the LogHub repository. The evaluation confirms that our approach achieves parsing accuracy on par with significantly larger models, such as LLaMA 70B and Anthropic's Claude. We further validate the practical utility of our fine-tuned LLM model by parsing over 600 million production logs from the Frontier supercomputer over a four-week window, uncovering critical patterns in temporal dynamics, node-level anomalies, and workload-error log correlations.
Fonte: arXiv cs.AI
RL • Score 85
OmniDiagram: Advancing Unified Diagram Code Generation via Visual Interrogation Reward
arXiv:2604.05514v1 Announce Type: new
Abstract: The paradigm of programmable diagram generation is evolving rapidly, playing a crucial role in structured visualization. However, most existing studies are confined to a narrow range of task formulations and language support, constraining their applicability to diverse diagram types. In this work, we propose OmniDiagram, a unified framework that incorporates diverse diagram code languages and task definitions. To address the challenge of aligning code logic with visual fidelity in Reinforcement Learning (RL), we introduce a novel visual feedback strategy named Visual Interrogation Verifies All (\textsc{Viva}). Unlike brittle syntax-based rules or pixel-level matching, \textsc{Viva} rewards the visual structure of rendered diagrams through a generative approach. Specifically, \textsc{Viva} actively generates targeted visual inquiries to scrutinize diagram visual fidelity and provides fine-grained feedback for optimization. This mechanism facilitates a self-evolving training process, effectively obviating the need for manually annotated ground truth code. Furthermore, we construct M3$^2$Diagram, the first large-scale diagram code generation dataset, containing over 196k high-quality instances. Experimental results confirm that the combination of SFT and our \textsc{Viva}-based RL allows OmniDiagram to establish a new state-of-the-art (SOTA) across diagram code generation benchmarks.
Fonte: arXiv cs.AI
RL • Score 85
TRACE: Capability-Targeted Agentic Training
arXiv:2604.05336v1 Announce Type: new
Abstract: Large Language Models (LLMs) deployed in agentic environments must exercise multiple capabilities across different task instances, where a capability is performing one or more actions in a trajectory that are necessary for successfully solving a subset of tasks in the environment. Many existing approaches either rely on synthetic training data that is not targeted to the model's actual capability deficits in the target environment or train directly on the target environment, where the model needs to implicitly learn the capabilities across tasks. We introduce TRACE (Turning Recurrent Agent failures into Capability-targeted training Environments), an end-to-end system for environment-specific agent self-improvement. TRACE contrasts successful and failed trajectories to automatically identify lacking capabilities, synthesizes a targeted training environment for each that rewards whether the capability was exercised, and trains a LoRA adapter via RL on each synthetic environment, routing to the relevant adapter at inference. Empirically, TRACE generalizes across different environments, improving over the base agent by +14.1 points on $\tau^2$-bench (customer service) and +7 perfect scores on ToolSandbox (tool use), outperforming the strongest baseline by +7.4 points and +4 perfect scores, respectively. Given the same number of rollouts, TRACE scales more efficiently than baselines, outperforming GRPO and GEPA by +9.2 and +7.4 points on $\tau^2$-bench.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
PaperOrchestra: A Multi-Agent Framework for Automated AI Research Paper Writing
arXiv:2604.05018v1 Announce Type: new
Abstract: Synthesizing unstructured research materials into manuscripts is an essential yet under-explored challenge in AI-driven scientific discovery. Existing autonomous writers are rigidly coupled to specific experimental pipelines, and produce superficial literature reviews. We introduce PaperOrchestra, a multi-agent framework for automated AI research paper writing. It flexibly transforms unconstrained pre-writing materials into submission-ready LaTeX manuscripts, including comprehensive literature synthesis and generated visuals, such as plots and conceptual diagrams. To evaluate performance, we present PaperWritingBench, the first standardized benchmark of reverse-engineered raw materials from 200 top-tier AI conference papers, alongside a comprehensive suite of automated evaluators. In side-by-side human evaluations, PaperOrchestra significantly outperforms autonomous baselines, achieving an absolute win rate margin of 50%-68% in literature review quality, and 14%-38% in overall manuscript quality.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Adaptive Serverless Resource Management via Slot-Survival Prediction and Event-Driven Lifecycle Control
arXiv:2604.05465v1 Announce Type: new
Abstract: Serverless computing eliminates infrastructure management overhead but introduces significant challenges regarding cold start latency and resource utilization. Traditional static resource allocation often leads to inefficiencies under variable workloads, resulting in performance degradation or excessive costs. This paper presents an adaptive engineering framework that optimizes serverless performance through event-driven architecture and probabilistic modeling. We propose a dual-strategy mechanism that dynamically adjusts idle durations and employs an intelligent request waiting strategy based on slot survival predictions. By leveraging sliding window aggregation and asynchronous processing, our system proactively manages resource lifecycles. Experimental results show that our approach reduces cold starts by up to 51.2% and improves cost-efficiency by nearly 2x compared to baseline methods in multi-cloud environments.
Fonte: arXiv cs.AI
RL • Score 85
Not All Agents Matter: From Global Attention Dilution to Risk-Prioritized Game Planning
arXiv:2604.05449v1 Announce Type: new
Abstract: End-to-end autonomous driving resides not in the integration of perception and planning, but rather in the dynamic multi-agent game within a unified representation space. Most existing end-to-end models treat all agents equally, hindering the decoupling of real collision threats from complex backgrounds. To address this issue, We introduce the concept of Risk-Prioritized Game Planning, and propose GameAD, a novel framework that models end-to-end autonomous driving as a risk-aware game problem. The GameAD integrates Risk-Aware Topology Anchoring, Strategic Payload Adapter, Minimax Risk-Aware Sparse Attention, and Risk Consistent Equilibrium Stabilization to enable game theoretic decision making with risk prioritized interactions. We also present the Planning Risk Exposure metric, which quantifies the cumulative risk intensity of planned trajectories over a long horizon for safe autonomous driving. Extensive experiments on the nuScenes and Bench2Drive datasets show that our approach significantly outperforms state-of-the-art methods, especially in terms of trajectory safety.
Fonte: arXiv cs.CV
Privacy/Security/Fairness • Score 85
Efficient machine unlearning with minimax optimality
arXiv:2604.05669v1 Announce Type: new
Abstract: There is a growing demand for efficient data removal to comply with regulations like the GDPR and to mitigate the influence of biased or corrupted data. This has motivated the field of machine unlearning, which aims to eliminate the influence of specific data subsets without the cost of full retraining. In this work, we propose a statistical framework for machine unlearning with generic loss functions and establish theoretical guarantees. For squared loss, especially, we develop Unlearning Least Squares (ULS) and establish its minimax optimality for estimating the model parameter of remaining data when only the pre-trained estimator, forget samples, and a small subsample of the remaining data are available. Our results reveal that the estimation error decomposes into an oracle term and an unlearning cost determined by the forget proportion and the forget model bias. We further establish asymptotically valid inference procedures without requiring full retraining. Numerical experiments and real-data applications demonstrate that the proposed method achieves performance close to retraining while requiring substantially less data access.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
HYVE: Hybrid Views for LLM Context Engineering over Machine Data
arXiv:2604.05400v1 Announce Type: new
Abstract: Machine data is central to observability and diagnosis in modern computing systems, appearing in logs, metrics, telemetry traces, and configuration snapshots. When provided to large language models (LLMs), this data typically arrives as a mixture of natural language and structured payloads such as JSON or Python/AST literals. Yet LLMs remain brittle on such inputs, particularly when they are long, deeply nested, and dominated by repetitive structure.
We present HYVE (HYbrid ViEw), a framework for LLM context engineering for inputs containing large machine-data payloads, inspired by database management principles. HYVE surrounds model invocation with coordinated preprocessing and postprocessing, centered on a request-scoped datastore augmented with schema information. During preprocessing, HYVE detects repetitive structure in raw inputs, materializes it in the datastore, transforms it into hybrid columnar and row-oriented views, and selectively exposes only the most relevant representation to the LLM. During postprocessing, HYVE either returns the model output directly, queries the datastore to recover omitted information, or performs a bounded additional LLM call for SQL-augmented semantic synthesis.
We evaluate HYVE on diverse real-world workloads spanning knowledge QA, chart generation, anomaly detection, and multi-step network troubleshooting. Across these benchmarks, HYVE reduces token usage by 50-90% while maintaining or improving output quality. On structured generation tasks, it improves chart-generation accuracy by up to 132% and reduces latency by up to 83%. Overall, HYVE offers a practical approximation to an effectively unbounded context window for prompts dominated by large machine-data payloads.
Fonte: arXiv cs.AI
Vision • Score 90
RCP: Representation Consistency Pruner for Mitigating Distribution Shift in Large Vision-Language Models
arXiv:2604.04972v1 Announce Type: new
Abstract: Large Vision-Language Models (LVLMs) suffer from prohibitive inference costs due to the massive number of visual tokens processed by the language decoder. Existing pruning methods often lead to significant performance degradation because the irreversible removal of visual tokens causes a distribution shift in the hidden states that deviates from the pre-trained full-token regime. To address this, we propose Representation Consistency Pruner, which we refer to as RCP, as a novel framework that integrates cumulative visual token pruning with a delayed repair mechanism. Specifically, we introduce a cross-attention pruner that leverages the intrinsic attention of the LLM as a baseline to predict cumulative masks, ensuring consistent and monotonic token reduction across layers. To compensate for the resulting information loss, we design a delayed repair adapter denoted as DRA, which caches the essence of pruned tokens and applies FiLM-based modulation specifically to the answer generation tokens. We employ a repair loss to match the first and second-order statistics of the pruned representations with a full-token teacher. RCP is highly efficient because it trains only lightweight plug-in modules while allowing for physical token discarding at inference. Extensive experiments on LVLM benchmarks demonstrate that RCP removes up to 88.9\% of visual tokens and reduces FLOPs by up to 85.7\% with only a marginal average accuracy drop, and outperforms prior methods that avoid fine-tuning the original model on several widely used benchmarks.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Don't Act Blindly: Robust GUI Automation via Action-Effect Verification and Self-Correction
arXiv:2604.05477v1 Announce Type: new
Abstract: Autonomous GUI agents based on vision-language models (VLMs) often assume deterministic environment responses, generating actions without verifying whether previous operations succeeded. In real-world settings with network latency, rendering delays, and system interruptions, this assumption leads to undetected action failures, repetitive ineffective behaviors, and catastrophic error accumulation. Moreover, learning robust recovery strategies is challenging due to the high cost of online interaction and the lack of real-time feedback in offline datasets.We propose VeriGUI (Verification-driven GUI Agent), which explicitly models action outcomes and recovery under noisy environments. VeriGUI introduces a Thinking--Verification--Action--Expectation (TVAE) framework to detect failures and guide corrective reasoning, and a two-stage training pipeline that combines Robust SFT with synthetic failure trajectories and GRPO with asymmetric verification rewards. We further construct a Robustness Benchmark based on AndroidControl to evaluate failure recognition and correction. Experiments show that VeriGUI significantly reduces failure loops and improves recovery success while maintaining competitive standard task performance.
Fonte: arXiv cs.CL
MLOps/Systems • Score 85
Triplet Feature Fusion for Equipment Anomaly Prediction : An Open-Source Methodology Using Small Foundation Models
arXiv:2602.15089v2 Announce Type: replace-cross
Abstract: Predicting equipment anomalies before they escalate into failures is a critical challenge in industrial facility management. Existing approaches rely either on hand-crafted threshold rules, which lack generalizability, or on large neural models that are impractical for on-site, air-gapped deployments. We present an industrial methodology that resolves this tension by combining open-source small foundation models into a unified 1,116-dimensional Triplet Feature Fusion pipeline. This pipeline integrates: (1) statistical features (x in $R^{28}$) derived from 90-day sensor histories, (2) time-series embeddings (y in $R^{64}$) from a LoRA-adapted IBM Granite TinyTimeMixer (TTM, 133K parameters), and (3) multilingual text embeddings (z in $R^{1024}$) extracted from Japanese equipment master records via multilingual-e5-large. The concatenated triplet h = [x; y; z] is processed by a LightGBM classifier (< 3 MB) trained to predict anomalies at 30-, 60-, and 90-day horizons. All components use permissive open-source licenses (Apache 2.0 / MIT). The inference-time pipeline runs entirely on CPU in under 2 ms, enabling edge deployment on co-located hardware without cloud dependency. On a dataset of 64 HVAC units comprising 67,045 samples, the triplet model achieves Precision = 0.992, F1 = 0.958, and ROC-AUC = 0.998 at the 30-day horizon. Crucially, it reduces the False Positive Rate from 0.6 percent (baseline) to 0.1 percent - an 83 percent reduction attributable to equipment-type conditioning via text embedding z. Cluster analysis reveals that the embeddings align time-series signatures with distinct fault archetypes, explaining how compact multilingual representations improve discrimination without explicit categorical encoding.
Fonte: arXiv stat.ML
MLOps/Systems • Score 90
AutoSOTA: An End-to-End Automated Research System for State-of-the-Art AI Model Discovery
arXiv:2604.05550v1 Announce Type: new
Abstract: Artificial intelligence research increasingly depends on prolonged cycles of reproduction, debugging, and iterative refinement to achieve State-Of-The-Art (SOTA) performance, creating a growing need for systems that can accelerate the full pipeline of empirical model optimization. In this work, we introduce AutoSOTA, an end-to-end automated research system that advances the latest SOTA models published in top-tier AI papers to reproducible and empirically improved new SOTA models. We formulate this problem through three tightly coupled stages: resource preparation and goal setting; experiment evaluation; and reflection and ideation. To tackle this problem, AutoSOTA adopts a multi-agent architecture with eight specialized agents that collaboratively ground papers to code and dependencies, initialize and repair execution environments, track long-horizon experiments, generate and schedule optimization ideas, and supervise validity to avoid spurious gains. We evaluate AutoSOTA on recent research papers collected from eight top-tier AI conferences under filters for code availability and execution cost. Across these papers, AutoSOTA achieves strong end-to-end performance in both automated replication and subsequent optimization. Specifically, it successfully discovers 105 new SOTA models that surpass the original reported methods, averaging approximately five hours per paper. Case studies spanning LLM, NLP, computer vision, time series, and optimization further show that the system can move beyond routine hyperparameter tuning to identify architectural innovation, algorithmic redesigns, and workflow-level improvements. These results suggest that end-to-end research automation can serve not only as a performance optimizer, but also as a new form of research infrastructure that reduces repetitive experimental burden and helps redirect human attention toward higher-level scientific creativity.
Fonte: arXiv cs.CL
MLOps/Systems • Score 85
fastml: Guarded Resampling Workflows for Safer Automated Machine Learning in R
arXiv:2604.05225v1 Announce Type: cross
Abstract: Preprocessing leakage arises when scaling, imputation, or other data-dependent transformations are estimated before resampling, inflating apparent performance while remaining hard to detect. We present fastml, an R package that provides a single-call interface for leakage-aware machine learning through guarded resampling, where preprocessing is re-estimated inside each resample and applied to the corresponding assessment data. The package supports grouped and time-ordered resampling, blocks high-risk configurations, audits recipes for external dependencies, and includes sandboxed execution and integrated model explanation. We evaluate fastml with a Monte Carlo simulation contrasting global and fold-local normalization, a usability comparison with tidymodels under matched specifications, and survival benchmarks across datasets of different sizes. The simulation demonstrates that global preprocessing substantially inflates apparent performance relative to guarded resampling. fastml matched held-out performance obtained with tidymodels while reducing workflow orchestration, and it supported consistent benchmarking of multiple survival model classes through a unified interface.
Fonte: arXiv stat.ML
NLP/LLMs • Score 90
See the Forest for the Trees: Loosely Speculative Decoding via Visual-Semantic Guidance for Efficient Inference of Video LLMs
arXiv:2604.05650v1 Announce Type: new
Abstract: Video Large Language Models (Video-LLMs) excel in video understanding but suffer from high inference latency during autoregressive generation. Speculative Decoding (SD) mitigates this by applying a draft-and-verify paradigm, yet existing methods are constrained by rigid exact-match rules, severely limiting the acceleration potential. To bridge this gap, we propose LVSpec, the first training-free loosely SD framework tailored for Video-LLMs. Grounded in the insight that generation is governed by sparse visual-relevant anchors (mandating strictness) amidst abundant visual-irrelevant fillers (permitting loose verification), LVSpec employs a lightweight visual-relevant token identification scheme to accurately pinpoint the former. To further maximize acceptance, we augment this with a position-shift tolerant mechanism that effectively salvages positionally mismatched but semantically equivalent tokens. Experiments demonstrate that LVSpec achieves high fidelity and speed: it preserves >99.8 of target performance while accelerating Qwen2.5-VL-32B by 2.70x and LLaVA-OneVision-72B by 2.94x. Notably, it boosts the mean accepted length and speedup ratio by 136% and 35% compared to SOTA training-free SD methods for Video-LLMs.
Fonte: arXiv cs.CL
Privacy/Security/Fairness • Score 85
From Governance Norms to Enforceable Controls: A Layered Translation Method for Runtime Guardrails in Agentic AI
arXiv:2604.05229v1 Announce Type: new
Abstract: Agentic AI systems plan, use tools, maintain state, and produce multi-step trajectories with external effects. Those properties create a governance problem that differs materially from single-turn generative AI: important risks emerge dur- ing execution, not only at model development or deployment time. Governance standards such as ISO/IEC 42001, ISO/IEC 23894, ISO/IEC 42005, ISO/IEC 5338, ISO/IEC 38507, and the NIST AI Risk Management Framework are therefore highly relevant to agentic AI, but they do not by themselves yield implementable runtime guardrails. This paper proposes a layered translation method that connects standards-derived governance objectives to four control layers: governance objectives, design- time constraints, runtime mediation, and assurance feedback. It distinguishes governance objectives, technical controls, runtime guardrails, and assurance evidence; introduces a control tuple and runtime-enforceability rubric for layer assignment; and demonstrates the method in a procurement-agent case study. The central claim is modest: standards should guide control placement across architecture, runtime policy, human escalation, and audit, while runtime guardrails are reserved for controls that are observable, determinate, and time-sensitive enough to justify execution-time intervention.
Fonte: arXiv cs.AI
Privacy/Security/Fairness • Score 85
RESIST: Resilient Decentralized Learning Using Consensus Gradient Descent
arXiv:2502.07977v2 Announce Type: replace-cross
Abstract: Empirical risk minimization (ERM) is a cornerstone of modern machine learning (ML), supported by advances in optimization theory that ensure efficient solutions with provable algorithmic and statistical learning rates. Privacy, memory, computation, and communication constraints necessitate data collection, processing, and storage across network-connected devices. In many applications, networks operate in decentralized settings where a central server cannot be assumed, requiring decentralized ML algorithms that are efficient and resilient. Decentralized learning, however, faces significant challenges, including an increased attack surface. This paper focuses on the man-in-the-middle (MITM) attack, wherein adversaries exploit communication vulnerabilities to inject malicious updates during training, potentially causing models to deviate from their intended ERM solutions. To address this challenge, we propose RESIST (Resilient dEcentralized learning using conSensus gradIent deScenT), an optimization algorithm designed to be robust against adversarially compromised communication links, where transmitted information may be arbitrarily altered before being received. Unlike existing adversarially robust decentralized learning methods, which often (i) guarantee convergence only to a neighborhood of the solution, (ii) lack guarantees of linear convergence for strongly convex problems, or (iii) fail to ensure statistical consistency as sample sizes grow, RESIST overcomes all three limitations. It achieves algorithmic and statistical convergence for strongly convex, Polyak-Lojasiewicz, and nonconvex ERM problems by employing a multistep consensus gradient descent framework and robust statistics-based screening methods to mitigate the impact of MITM attacks. Experimental results demonstrate the robustness and scalability of RESIST across attack strategies, screening methods, and loss functions.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
RUQuant: Towards Refining Uniform Quantization for Large Language Models
arXiv:2604.04013v1 Announce Type: new
Abstract: The increasing size and complexity of large language models (LLMs) have raised significant challenges in deployment efficiency, particularly under resource constraints. Post-training quantization (PTQ) has emerged as a practical solution by compressing models without requiring retraining. While existing methods focus on uniform quantization schemes for both weights and activations, they often suffer from substantial accuracy degradation due to the non-uniform nature of activation distributions. In this work, we revisit the activation quantization problem from a theoretical perspective grounded in the Lloyd-Max optimality conditions. We identify the core issue as the non-uniform distribution of activations within the quantization interval, which causes the optimal quantization point under the Lloyd-Max criterion to shift away from the midpoint of the interval. To address this issue, we propose a two-stage orthogonal transformation method, RUQuant. In the first stage, activations are divided into blocks. Each block is mapped to uniformly sampled target vectors using composite orthogonal matrices, which are constructed from Householder reflections and Givens rotations. In the second stage, a global Householder reflection is fine-tuned to further minimize quantization error using Transformer output discrepancies. Empirical results show that our method achieves near-optimal quantization performance without requiring model fine-tuning: RUQuant achieves 99.8% of full-precision accuracy with W6A6 and 97% with W4A4 quantization for a 13B LLM, within approximately one minute. A fine-tuned variant yields even higher accuracy, demonstrating the effectiveness and scalability of our approach.
Fonte: arXiv cs.CL
RecSys • Score 85
AdaptFuse: Training-Free Sequential Preference Learning via Externalized Bayesian Inference
arXiv:2604.03925v1 Announce Type: new
Abstract: Large language models struggle to accumulate evidence across multiple rounds of user interaction, failing to update their beliefs in a manner consistent with Bayesian inference. Existing solutions require fine-tuning on sensitive user interaction data, limiting their applicability in privacy-conscious settings. We propose AdaptFuse, a training-free framework that externalizes probabilistic computation entirely from the LLM: a symbolic module maintains a Bayesian posterior over a discrete hypothesis set, while a frozen LLM contributes semantic reasoning via multi-sample Dirichlet aggregation. The two signals are combined through entropy-adaptive fusion, which automatically weights each source by its predictive confidence, shifting reliance from the LLM to the symbolic posterior as evidence accumulates. We evaluate across three domains: flight recommendation, hotel recommendation, and web shopping; on Gemma 2 9B, Llama 3 8B, and Qwen 2.5 7B. AdaptFuse consistently outperforms both prompting baselines and fine-tuned Bayesian Teaching models on all tasks, with accuracy improving monotonically over interaction rounds. These results demonstrate that principled inference-time algorithms can substitute for fine-tuning in personalized recommendation, without storing or training on sensitive user data. All the code and materials will be open-sourced.
Fonte: arXiv cs.CL
Privacy/Security/Fairness • Score 85
Your Agent is More Brittle Than You Think: Uncovering Indirect Injection Vulnerabilities in Agentic LLMs
arXiv:2604.03870v1 Announce Type: new
Abstract: The rapid deployment of open-source frameworks has significantly advanced the development of modern multi-agent systems. However, expanded action spaces, including uncontrolled privilege exposure and hidden inter-system interactions, pose severe security challenges. Specifically, Indirect Prompt Injections (IPI), which conceal malicious instructions within third-party content, can trigger unauthorized actions such as data exfiltration during normal operations. While current security evaluations predominantly rely on isolated single-turn benchmarks, the systemic vulnerabilities of these agents within complex dynamic environments remain critically underexplored. To bridge this gap, we systematically evaluate six defense strategies against four sophisticated IPI attack vectors across nine LLM backbones. Crucially, we conduct our evaluation entirely within dynamic multi-step tool-calling environments to capture the true attack surface of modern autonomous agents. Moving beyond binary success rates, our multidimensional analysis reveals a pronounced fragility. Advanced injections successfully bypass nearly all baseline defenses, and some surface-level mitigations even produce counterproductive side effects. Furthermore, while agents execute malicious instructions almost instantaneously, their internal states exhibit abnormally high decision entropy. Motivated by this latent hesitation, we investigate Representation Engineering (RepE) as a robust detection strategy. By extracting hidden states at the tool-input position, we revealed that the RepE-based circuit breaker successfully identifies and intercepts unauthorized actions before the agent commits to them, achieving high detection accuracy across diverse LLM backbones. This study exposes the limitations of current IPI defenses and provides a highly practical paradigm for building resilient multi-agent architectures.
Fonte: arXiv cs.CL
Theory/Optimization • Score 85
A Muon-Accelerated Algorithm for Low Separation Rank Tensor Generalized Linear Models
arXiv:2604.04726v1 Announce Type: new
Abstract: Tensor-valued data arise naturally in multidimensional signal and imaging problems, such as biomedical imaging. When incorporated into generalized linear models (GLMs), naive vectorization can destroy their multi-way structure and lead to high-dimensional, ill-posed estimation. To address this challenge, Low Separation Rank (LSR) decompositions reduce model complexity by imposing low-rank multilinear structure on the coefficient tensor. A representative approach for estimating LSR-based tensor GLMs (LSR-TGLMs) is the Low Separation Rank Tensor Regression (LSRTR) algorithm, which adopts block coordinate descent and enforces orthogonality of the factor matrices through repeated QR-based projections. However, the repeated projection steps can be computationally demanding and slow convergence. Motivated by the need for scalable estimation and classification from such data, we propose LSRTR-M, which incorporates Muon (MomentUm Orthogonalized by Newton-Schulz) updates into the LSRTR framework. Specifically, LSRTR-M preserves the original block coordinate scheme while replacing the projection-based factor updates with Muon steps. Across synthetic linear, logistic, and Poisson LSR-TGLMs, LSRTR-M converges faster in both iteration count and wall-clock time, while achieving lower normalized estimation and prediction errors. On the Vessel MNIST 3D task, it further improves computational efficiency while maintaining competitive classification performance.
Fonte: arXiv stat.ML
Theory/Optimization • Score 85
The Geometric Alignment Tax: Tokenization vs. Continuous Geometry in Scientific Foundation Models
arXiv:2604.04155v1 Announce Type: cross
Abstract: Foundation models for biology and physics optimize predictive accuracy, but their internal representations systematically fail to preserve the continuous geometry of the systems they model. We identify the root cause: the Geometric Alignment Tax, an intrinsic cost of forcing continuous manifolds through discrete categorical bottlenecks. Controlled ablations on synthetic dynamical systems demonstrate that replacing cross-entropy with a continuous head on an identical encoder reduces geometric distortion by up to 8.5x, while learned codebooks exhibit a non-monotonic double bind where finer quantization worsens geometry despite improving reconstruction. Under continuous objectives, three architectures differ by 1.3x; under discrete tokenization, they diverge by 3,000x. Evaluating 14 biological foundation models with rate-distortion theory and MINE, we identify three failure regimes: Local-Global Decoupling, Representational Compression, and Geometric Vacuity. A controlled experiment confirms that Evo 2's reverse-complement robustness on real DNA reflects conserved sequence composition, not learned symmetry. No model achieves simultaneously low distortion, high mutual information, and global coherence.
Fonte: arXiv stat.ML
Theory/Optimization • Score 85
Importance Sparsification for Sinkhorn Algorithm
arXiv:2306.06581v2 Announce Type: replace
Abstract: Sinkhorn algorithm has been used pervasively to approximate the solution to optimal transport (OT) and unbalanced optimal transport (UOT) problems. However, its practical application is limited due to the high computational complexity. To alleviate the computational burden, we propose a novel importance sparsification method, called Spar-Sink, to efficiently approximate entropy-regularized OT and UOT solutions. Specifically, our method employs natural upper bounds for unknown optimal transport plans to establish effective sampling probabilities, and constructs a sparse kernel matrix to accelerate Sinkhorn iterations, reducing the computational cost of each iteration from $O(n^2)$ to $\widetilde{O}(n)$ for a sample of size $n$. Theoretically, we show the proposed estimators for the regularized OT and UOT problems are consistent under mild regularity conditions. Experiments on various synthetic data demonstrate Spar-Sink outperforms mainstream competitors in terms of both estimation error and speed. A real-world echocardiogram data analysis shows Spar-Sink can effectively estimate and visualize cardiac cycles, from which one can identify heart failure and arrhythmia. To evaluate the numerical accuracy of cardiac cycle prediction, we consider the task of predicting the end-systole time point using the end-diastole one. Results show Spar-Sink performs as well as the classical Sinkhorn algorithm, requiring significantly less computational time.
Fonte: arXiv stat.ML
Theory/Optimization • Score 85
An Analytical Theory of Spectral Bias in the Learning Dynamics of Diffusion Models
arXiv:2503.03206v3 Announce Type: replace-cross
Abstract: We develop an analytical framework for understanding how the generated distribution evolves during diffusion model training. Leveraging a Gaussian-equivalence principle, we solve the full-batch gradient-flow dynamics of linear and convolutional denoisers and integrate the resulting probability-flow ODE, yielding analytic expressions for the generated distribution. The theory exposes a universal inverse-variance spectral law: the time for an eigen- or Fourier mode to match its target variance scales as $\tau\propto\lambda^{-1}$, so high-variance (coarse) structure is mastered orders of magnitude sooner than low-variance (fine) detail. Extending the analysis to deep linear networks and circulant full-width convolutions shows that weight sharing merely multiplies learning rates -- accelerating but not eliminating the bias -- whereas local convolution introduces a qualitatively different bias. Experiments on Gaussian and natural-image datasets confirm the spectral law persists in deep MLP-based UNet. Convolutional U-Nets, however, display rapid near-simultaneous emergence of many modes, implicating local convolution in reshaping learning dynamics. These results underscore how data covariance governs the order and speed with which diffusion models learn, and they call for deeper investigation of the unique inductive biases introduced by local convolution.
Fonte: arXiv stat.ML
Multimodal • Score 85
CoLA: Cross-Modal Low-rank Adaptation for Multimodal Downstream Tasks
arXiv:2604.03314v1 Announce Type: new
Abstract: Foundation models have revolutionized AI, but adapting them efficiently for multimodal tasks, particularly in dual-stream architectures composed of unimodal encoders, such as DINO and BERT, remains a significant challenge. Parameter-Efficient Fine-Tuning (PEFT) methods like Low-Rank Adaptation (LoRA) enable lightweight adaptation, yet they operate in isolation within each modality, limiting their ability in capturing cross-modal interactions. In this paper, we take a step in bridging this gap with Cross-Modal Low-Rank Adaptation (CoLA), a novel PEFT framework that extends LoRA by introducing a dedicated inter-modal adaptation pathway alongside the standard intra-modal one. This dual-path design enables CoLA to adapt unimodal foundation models to multimodal tasks effectively, without interference between modality-specific and cross-modal learning. We evaluate CoLA across a range of vision-language (RefCOCO, RefCOCO+, RefCOCOg) and audio-visual (AVE, AVS) benchmarks, where it consistently outperforms LORA, achieving a relative gain of around 3\% and 2\%, respectively, while maintaining parameter efficiency. Notably, CoLA enables the first multi-task PEFT framework for visual grounding, bridging a key gap in efficient multimodal adaptation.
Fonte: arXiv cs.CV
Theory/Optimization • Score 85
Hardware-Oriented Inference Complexity of Kolmogorov-Arnold Networks
arXiv:2604.03345v1 Announce Type: new
Abstract: Kolmogorov-Arnold Networks (KANs) have recently emerged as a powerful architecture for various machine learning applications. However, their unique structure raises significant concerns regarding their computational overhead. Existing studies primarily evaluate KAN complexity in terms of Floating-Point Operations (FLOPs) required for GPU-based training and inference. However, in many latency-sensitive and power-constrained deployment scenarios, such as neural network-driven non-linearity mitigation in optical communications or channel state estimation in wireless communications, training is performed offline and dedicated hardware accelerators are preferred over GPUs for inference. Recent hardware implementation studies report KAN complexity using platform-specific resource consumption metrics, such as Look-Up Tables, Flip-Flops, and Block RAMs. However, these metrics require a full hardware design and synthesis stage that limits their utility for early-stage architectural decisions and cross-platform comparisons. To address this, we derive generalized, platform-independent formulae for evaluating the hardware inference complexity of KANs in terms of Real Multiplications (RM), Bit Operations (BOP), and Number of Additions and Bit-Shifts (NABS). We extend our analysis across multiple KAN variants, including B-spline, Gaussian Radial Basis Function (GRBF), Chebyshev, and Fourier KANs. The proposed metrics can be computed directly from the network structure and enable a fair and straightforward inference complexity comparison between KAN and other neural network architectures.
Fonte: arXiv cs.LG
Multimodal • Score 85
Multimodal Urban Tree Detection from Satellite and Street-Level Imagery via Annotation-Efficient Deep Learning Strategies
arXiv:2604.03505v1 Announce Type: new
Abstract: Beyond the immediate biophysical benefits, urban trees play a foundational role in environmental sustainability and disaster mitigation. Precise mapping of urban trees is essential for environmental monitoring, post-disaster assessment, and strengthening policy. However, the transition from traditional, labor-intensive field surveys to scalable automated systems remains limited by high annotation costs and poor generalization across diverse urban scenarios. This study introduces a multimodal framework that integrates high-resolution satellite imagery with ground-level Google Street View to enable scalable and detailed urban tree detection under limited-annotation conditions. The framework first leverages satellite imagery to localize tree candidates and then retrieves targeted ground-level views for detailed detection, significantly reducing inefficient street-level sampling. To address the annotation bottleneck, domain adaptation is used to transfer knowledge from an existing annotated dataset to a new region of interest. To further minimize human effort, we evaluated three learning strategies: semi-supervised learning, active learning, and a hybrid approach combining both, using a transformer-based detection model. The hybrid strategy achieved the best performance with an F1-score of 0.90, representing a 12% improvement over the baseline model. In contrast, semi-supervised learning exhibited progressive performance degradation due to confirmation bias in pseudo-labeling, while active learning steadily improved results through targeted human intervention to label uncertain or incorrect predictions. Error analysis further showed that active and hybrid strategies reduced both false positives and false negatives. Our findings highlight the importance of a multimodal approach and guided annotation for scalable, annotation-efficient urban tree mapping to strengthen sustainable city planning.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
SoLA: Leveraging Soft Activation Sparsity and Low-Rank Decomposition for Large Language Model Compression
arXiv:2604.03258v1 Announce Type: new
Abstract: Large language models (LLMs) have demonstrated impressive capabilities across various tasks, but the billion-scale parameters pose deployment challenges. Although existing methods attempt to reduce the scale of LLMs, they require either special hardware support or expensive post-training to maintain model quality. To facilitate efficient and affordable model slimming, we propose a novel training-free compression method for LLMs, named "SoLA", which leverages \textbf{So}ft activation sparsity and \textbf{L}ow-r\textbf{A}nk decomposition. SoLA can identify and retain a minority of components significantly contributing to inference, while compressing the majority through low-rank decomposition, based on our analysis of the activation pattern in the feed-forward network (FFN) of modern LLMs. To alleviate the decomposition loss, SoLA is equipped with an adaptive component-wise low-rank allocation strategy to assign appropriate truncation positions for different weight matrices. We conduct extensive experiments on LLaMA-2-7B/13B/70B and Mistral-7B models across a variety of benchmarks. SoLA exhibits remarkable improvement in both language modeling and downstream task accuracy without post-training. For example, with a 30\% compression rate on the LLaMA-2-70B model, SoLA surpasses the state-of-the-art method by reducing perplexity from 6.95 to 4.44 and enhancing downstream task accuracy by 10\%.
Fonte: arXiv cs.CL
Vision • Score 85
Edge-Based Standing-Water Detection via FSM-Guided Tiering and Multi-Model Consensus
arXiv:2604.03308v1 Announce Type: new
Abstract: Standing water in agricultural fields threatens vehicle mobility and crop health. This paper presents a deployed edge architecture for standing-water detection using Raspberry-Pi-class devices with optional Jetson acceleration. Camera input and environmental sensors (humidity, pressure, temperature) are combined in a finite-state machine (FSM) that acts as the architectural decision engine. The FSM-guided control plane selects between local and offloaded inference tiers, trading accuracy, latency, and energy under intermittent connectivity and motion-dependent compute budgets. A multi-model YOLO ensemble provides image scores, while diurnal-baseline sensor fusion adjusts caution using environmental anomalies. All decisions are logged per frame, enabling bit-identical hardware-in-the-loop replays. Across ten configurations and sensor variants on identical field sequences with frame-level ground truth, we show that the combination of adaptive tiering, multi-model consensus, and diurnal sensor fusion improves flood-detection performance over static local baselines, uses less energy than a naive always-heavy offload policy, and maintains bounded tail latency in a real agricultural setting.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
From XAI to MLOps: Explainable Concept Drift Detection with Profile Drift Detection
arXiv:2412.11308v2 Announce Type: replace
Abstract: Predictive models often degrade in performance due to evolving data distributions, a phenomenon known as data drift. Among its forms, concept drift, where the relationship between explanatory variables and the response variable changes, is particularly challenging to detect and adapt to. Traditional drift detection methods often rely on metrics such as accuracy or marginal variable distributions, which may fail to capture subtle but important conceptual changes. This paper proposes a novel method, Profile Drift Detection (PDD), which enables both the detection of concept drift and an enhanced understanding of its underlying causes by leveraging an explainable AI tool: Partial Dependence Profiles (PDPs). PDD quantifies changes in PDPs through new drift metrics that are sensitive to shifts in the data stream while remaining computationally efficient. This approach is aligned with MLOps practices, emphasizing continuous model monitoring and adaptive retraining in dynamic environments. Experiments on synthetic and real-world datasets demonstrate that PDD outperforms existing methods by maintaining high predictive performance while effectively balancing sensitivity and stability in drift signals. The results highlight its suitability for real-time applications, and the paper concludes by discussing the method's advantages, limitations, and potential extensions to broader use cases.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
KiToke: Kernel-based Interval-aware Token Compression for Video Large Language Models
arXiv:2604.03414v1 Announce Type: new
Abstract: Video Large Language Models (Video LLMs) achieve strong performance on video understanding tasks but suffer from high inference costs due to the large number of visual tokens. We propose KiToke, a training-free, query-agnostic token compression approach that reduces spatiotemporal redundancy while preserving critical visual information. Our method estimates token diversity globally using a kernel-based redundancy measure, enabling content-adaptive selection that remains effective under extreme token budgets, and further introduces a lightweight temporal interval construction with interval-aware token merging to maintain temporal coherence. Unlike prior methods that rely on local or segment-level heuristics, KiToke explicitly captures global redundancy across an entire video, leading to more efficient token utilization. Extensive experiments on multiple video understanding benchmarks and Video LLM backbones demonstrate that KiToke consistently outperforms existing training-free compression methods, with particularly large gains at aggressive retention ratios down to 1%.
Fonte: arXiv cs.CV
Vision • Score 85
A reconfigurable smart camera implementation for jet flames characterization based on an optimized segmentation model
arXiv:2604.03267v1 Announce Type: new
Abstract: In this work we present a novel framework for fire safety management in industrial settings through the implementation of a smart camera platform for jet flames characterization. The approach seeks to alleviate the lack of real-time solutions for industrial early fire segmentation and characterization. As a case study, we demonstrate how a SoC FPGA, running optimized Artificial Intelligence (AI) models can be leveraged to implement a full edge processing pipeline for jet flames analysis. In this paper we extend previous work on computer-vision jet fire segmentation by creating a novel experimental set-up and system implementation for addressing this issue, which can be replicated to other fire safety applications. The proposed platform is designed to carry out image processing tasks in real-time and on device, reducing video processing overheads, and thus the overall latency. This is achieved by optimizing a UNet segmentation model to make it amenable for an SoC FPGAs implementation; the optimized model can then be efficiently mapped onto the SoC reconfigurable logic for massively parallel execution. For our experiments, we have chosen the Ultra96 platform, as it also provides the means for implementing full-fledged intelligent systems using the SoC peripherals, as well as other Operating System (OS) capabilities (i.e., multi-threading) for systems management. For optimizing the model we made use of the Vitis (Xilinx) framework, which enabled us to optimize the full precision model from 7.5 million parameters to 59,095 parameters (125x less), which translated into a reduction of the processing latency of 2.9x. Further optimization (multi-threading and batch normalization) led to an improvement of 7.5x in terms of latency, yielding a performance of 30 Frames Per Second (FPS) without sacrificing accuracy in terms of the evaluated metrics (Dice Score).
Fonte: arXiv cs.CV
Applications • Score 85
Integrating Artificial Intelligence, Physics, and Internet of Things: A Framework for Cultural Heritage Conservation
arXiv:2604.03233v1 Announce Type: new
Abstract: The conservation of cultural heritage increasingly relies on integrating technological innovation with domain expertise to ensure effective monitoring and predictive maintenance. This paper presents a novel framework to support the preservation of cultural assets, combining Internet of Things (IoT) and Artificial Intelligence (AI) technologies, enhanced with the physical knowledge of phenomena. The framework is structured into four functional layers that permit the analysis of 3D models of cultural assets and elaborate simulations based on the knowledge acquired from data and physics. A central component of the proposed framework consists of Scientific Machine Learning, particularly Physics-Informed Neural Networks (PINNs), which incorporate physical laws into deep learning models. To enhance computational efficiency, the framework also integrates Reduced Order Methods (ROMs), specifically Proper Orthogonal Decomposition (POD), and is also compatible with classical Finite Element (FE) methods. Additionally, it includes tools to automatically manage and process 3D digital replicas, enabling their direct use in simulations. The proposed approach offers three main contributions: a methodology for processing 3D models of cultural assets for reliable simulation; the application of PINNs to combine data-driven and physics-based approaches in cultural heritage conservation; and the integration of PINNs with ROMs to efficiently model degradation processes influenced by environmental and material parameters. The reproducible and open-access experimental phase exploits simulated scenarios on complex and real-life geometries to test the efficacy of the proposed framework in each of its key components, allowing the possibility of dealing with both direct and inverse problems. Code availability: https://github.com/valc89/PhysicsInformedCulturalHeritage
Fonte: arXiv cs.LG
Vision • Score 85
Automated Segmentation and Tracking of Group Housed Pigs Using Foundation Models
arXiv:2604.03426v1 Announce Type: new
Abstract: Foundation models (FM) are reshaping computer vision by reducing reliance on task-specific supervised learning and leveraging general visual representations learned at scale. In precision livestock farming, most pipelines remain dominated by supervised learning models that require extensive labeled data, repeated retraining, and farm-specific tuning. This study presents an FM-centered workflow for automated monitoring of group-housed nursery pigs, in which pretrained vision-language FM serve as general visual backbones and farm-specific adaptation is achieved through modular post-processing. Grounding-DINO was first applied to 1,418 annotated images to establish a baseline detection performance. While detection accuracy was high under daytime conditions, performance degraded under night-vision and heavy occlusion, motivating the integration of temporal tracking logic. Building on these detections, short-term video segmentation with Grounded-SAM2 was evaluated on 550 one-minute video clips; after post-processing, over 80% of 4,927 active tracks were fully correct, with most remaining errors arising from inaccurate masks or duplicated labels. To support identity consistency over an extended time, we further developed a long-term tracking pipeline integrating initialization, tracking, matching, mask refinement, re-identification, and post-hoc quality control. This system was evaluated on a continuous 132-minute video and maintained stable identities throughout. On 132 uniformly sampled ground-truth frames, the system achieved a mean region similarity (J) of 0.83, contour accuracy (F) of 0.92, J&F of 0.87, MOTA of 0.99, and MOTP of 90.7%, with no identity switches. Overall, this work demonstrates how FM prior knowledge can be combined with lightweight, task-specific logic to enable scalable, label-efficient, and long-duration monitoring in pig production.
Fonte: arXiv cs.CV
Theory/Optimization • Score 85
Improving Feasibility via Fast Autoencoder-Based Projections
arXiv:2604.03489v1 Announce Type: new
Abstract: Enforcing complex (e.g., nonconvex) operational constraints is a critical challenge in real-world learning and control systems. However, existing methods struggle to efficiently enforce general classes of constraints. To address this, we propose a novel data-driven amortized approach that uses a trained autoencoder as an approximate projector to provide fast corrections to infeasible predictions. Specifically, we train an autoencoder using an adversarial objective to learn a structured, convex latent representation of the feasible set. This enables rapid correction of neural network outputs by projecting their associated latent representations onto a simple convex shape before decoding into the original feasible set. We test our approach on a diverse suite of constrained optimization and reinforcement learning problems with challenging nonconvex constraints. Results show that our method effectively enforces constraints at a low computational cost, offering a practical alternative to expensive feasibility correction techniques based on traditional solvers.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
BlazeFL: Fast and Deterministic Federated Learning Simulation
arXiv:2604.03606v1 Announce Type: new
Abstract: Federated learning (FL) research increasingly relies on single-node simulations with hundreds or thousands of virtual clients, making both efficiency and reproducibility essential. Yet parallel client training often introduces nondeterminism through shared random state and scheduling variability, forcing researchers to trade throughput for reproducibility or to implement custom control logic within complex frameworks. We present BlazeFL, a lightweight framework for single-node FL simulation that alleviates this trade-off through free-threaded shared-memory execution and deterministic randomness management. BlazeFL uses thread-based parallelism with in-memory parameter exchange between the server and clients, avoiding serialization and inter-process communication overhead. To support deterministic execution, BlazeFL assigns isolated random number generator (RNG) streams to clients. Under a fixed software/hardware stack, and when stochastic operators consume BlazeFL-managed generators, this design yields bitwise-identical results across repeated high-concurrency runs in both thread-based and process-based modes. In CIFAR-10 image-classification experiments, BlazeFL substantially reduces execution time relative to a widely used open-source baseline, achieving up to 3.1$\times$ speedup on communication-dominated workloads while preserving a lightweight dependency footprint. Our open-source implementation is available at: https://github.com/kitsuyaazuma/blazefl.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
From Model-Based Screening to Data-Driven Surrogates: A Multi-Stage Workflow for Exploring Stochastic Agent-Based Models
arXiv:2604.03350v1 Announce Type: new
Abstract: Systematic exploration of Agent-Based Models (ABMs) is challenged by the curse of dimensionality and their inherent stochasticity. We present a multi-stage pipeline integrating the systematic design of experiments with machine learning surrogates. Using a predator-prey case study, our methodology proceeds in two steps. First, an automated model-based screening identifies dominant variables, assesses outcome variability, and segments the parameter space. Second, we train Machine Learning models to map the remaining nonlinear interaction effects. This approach automates the discovery of unstable regions where system outcomes are highly dependent on nonlinear interactions between many variables. Thus, this work provides modelers with a rigorous, hands-off framework for sensitivity analysis and policy testing, even when dealing with high-dimensional stochastic simulators.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Simple yet Effective: Low-Rank Spatial Attention for Neural Operators
arXiv:2604.03582v1 Announce Type: new
Abstract: Neural operators have emerged as data-driven surrogates for solving partial differential equations (PDEs), and their success hinges on efficiently modeling the long-range, global coupling among spatial points induced by the underlying physics. In many PDE regimes, the induced global interaction kernels are empirically compressible, exhibiting rapid spectral decay that admits low-rank approximations. We leverage this observation to unify representative global mixing modules in neural operators under a shared low-rank template: compressing high-dimensional pointwise features into a compact latent space, processing global interactions within it, and reconstructing the global context back to spatial points. Guided by this view, we introduce Low-Rank Spatial Attention (LRSA) as a clean and direct instantiation of this template. Crucially, unlike prior approaches that often rely on non-standard aggregation or normalization modules, LRSA is built purely from standard Transformer primitives, i.e., attention, normalization, and feed-forward networks, yielding a concise block that is straightforward to implement and directly compatible with hardware-optimized kernels. In our experiments, such a simple construction is sufficient to achieve high accuracy, yielding an average error reduction of over 17\% relative to second-best methods, while remaining stable and efficient in mixed-precision training.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Piecewise Deterministic Markov Processes for Bayesian Neural Networks
arXiv:2302.08724v4 Announce Type: replace
Abstract: Inference on modern Bayesian Neural Networks (BNNs) often relies on a variational inference treatment, imposing violated assumptions of independence and the form of the posterior. Traditional MCMC approaches avoid these assumptions at the cost of increased computation due to its incompatibility to subsampling of the likelihood. New Piecewise Deterministic Markov Process (PDMP) samplers permit subsampling, though introduce a model specific inhomogenous Poisson Process (IPPs) which is difficult to sample from. This work introduces a new generic and adaptive thinning scheme for sampling from these IPPs, and demonstrates how this approach can accelerate the application of PDMPs for inference in BNNs. Experimentation illustrates how inference with these methods is computationally feasible, can improve predictive accuracy, MCMC mixing performance, and provide informative uncertainty measurements when compared against other approximate inference schemes.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
DRAFT: Task Decoupled Latent Reasoning for Agent Safety
arXiv:2604.03242v1 Announce Type: new
Abstract: The advent of tool-using LLM agents shifts safety monitoring from output moderation to auditing long, noisy interaction trajectories, where risk-critical evidence is sparse-making standard binary supervision poorly suited for credit assignment. To address this, we propose DRAFT (Task Decoupled Latent Reasoning for Agent Safety), a latent reasoning framework that decouples safety judgment into two trainable stages: an Extractor that distills the full trajectory into a compact continuous latent draft, and a Reasoner that jointly attends to the draft and the original trajectory to predict safety. DRAFT avoids lossy explicit summarize-then-judge pipelines by performing evidence aggregation in latent space, enabling end-to-end differentiable training.Across benchmarks including ASSEBench and R-Judge, DRAFT consistently outperforms strong baselines, improving accuracy from 63.27% (LoRA) to 91.18% averaged over benchmarks, and learns more separable representations. Ablations demonstrate a clear synergy between the Extractor and the Reasoner.Overall, DRAFT suggests that continuous latent reasoning prior to readout is a practical path to robust agent safety under long-context supervision with sparse evidence.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
DARE: Diffusion Large Language Models Alignment and Reinforcement Executor
arXiv:2604.04215v1 Announce Type: new
Abstract: Diffusion large language models (dLLMs) are emerging as a compelling alternative to dominant autoregressive models, replacing strictly sequential token generation with iterative denoising and parallel generation dynamics. However, their open-source ecosystem remains fragmented across model families and, in particular, across post-training pipelines, where reinforcement learning objectives, rollout implementations and evaluation scripts are often released as paper-specific codebases. This fragmentation slows research iteration, raises the engineering burden of reproduction, and makes fair comparison across algorithms difficult. We present \textbf{DARE} (\textbf{d}LLMs \textbf{A}lignment and \textbf{R}einforcement \textbf{E}xecutor), an open framework for post-training and evaluating dLLMs. Built on top of verl~\cite{sheng2024hybridflow} and OpenCompass~\cite{2023opencompass}, DARE unifies supervised fine-tuning, parameter-efficient fine-tuning, preference optimization, and dLLM-specific reinforcement learning under a shared execution stack for both masked and block diffusion language models. Across representative model families including LLaDA, Dream, SDAR, and LLaDA2.x, DARE provides broad algorithmic coverage, reproducible benchmark evaluation, and practical acceleration. Extensive empirical results position that DARE serves as a reusable research substrate for developing, comparing, and deploying post-training methods for current and emerging dLLMs.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
Profile-Then-Reason: Bounded Semantic Complexity for Tool-Augmented Language Agents
arXiv:2604.04131v1 Announce Type: new
Abstract: Large language model agents that use external tools are often implemented through reactive execution, in which reasoning is repeatedly recomputed after each observation, increasing latency and sensitivity to error propagation. This work introduces Profile--Then--Reason (PTR), a bounded execution framework for structured tool-augmented reasoning, in which a language model first synthesizes an explicit workflow, deterministic or guarded operators execute that workflow, a verifier evaluates the resulting trace, and repair is invoked only when the original workflow is no longer reliable. A mathematical formulation is developed in which the full pipeline is expressed as a composition of profile, routing, execution, verification, repair, and reasoning operators; under bounded repair, the number of language-model calls is restricted to two in the nominal case and three in the worst case. Experiments against a ReAct baseline on six benchmarks and four language models show that PTR achieves the pairwise exact-match advantage in 16 of 24 configurations. The results indicate that PTR is particularly effective on retrieval-centered and decomposition-heavy tasks, whereas reactive execution remains preferable when success depends on substantial online adaptation.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
ActionNex: A Virtual Outage Manager for Cloud
arXiv:2604.03512v1 Announce Type: new
Abstract: Outage management in large-scale cloud operations remains heavily manual, requiring rapid triage, cross-team coordination, and experience-driven decisions under partial observability. We present \textbf{ActionNex}, a production-grade agentic system that supports end-to-end outage assistance, including real-time updates, knowledge distillation, and role- and stage-conditioned next-best action recommendations. ActionNex ingests multimodal operational signals (e.g., outage content, telemetry, and human communications) and compresses them into critical events that represent meaningful state transitions. It couples this perception layer with a hierarchical memory subsystem: long-term Key-Condition-Action (KCA) knowledge distilled from playbooks and historical executions, episodic memory of prior outages, and working memory of the live context. A reasoning agent aligns current critical events to preconditions, retrieves relevant memories, and generates actionable recommendations; executed human actions serve as an implicit feedback signal to enable continual self-evolution in a human-agent hybrid system. We evaluate ActionNex on eight real Azure outages (8M tokens, 4,000 critical events) using two complementary ground-truth action sets, achieving 71.4\% precision and 52.8-54.8\% recall. The system has been piloted in production and has received positive early feedback.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
Partially deterministic sampling for compressed sensing with denoising guarantees
arXiv:2604.04802v1 Announce Type: cross
Abstract: We study compressed sensing when the sampling vectors are chosen from the rows of a unitary matrix. In the literature, these sampling vectors are typically chosen randomly; the use of randomness has enabled major empirical and theoretical advances in the field. However, in practice there are often certain crucial sampling vectors, in which case practitioners will depart from the theory and sample such rows deterministically. In this work, we derive an optimized sampling scheme for Bernoulli selectors which naturally combines random and deterministic selection of rows, thus rigorously deciding which rows should be sampled deterministically. This sampling scheme provides measurable improvements in image compressed sensing for both generative and sparse priors when compared to with-replacement and without-replacement sampling schemes, as we show with theoretical results and numerical experiments. Additionally, our theoretical guarantees feature improved sample complexity bounds compared to previous works, and novel denoising guarantees in this setting.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Towards Intelligent Energy Security: A Unified Spatio-Temporal and Graph Learning Framework for Scalable Electricity Theft Detection in Smart Grids
arXiv:2604.03344v1 Announce Type: new
Abstract: Electricity theft and non-technical losses (NTLs) remain critical challenges in modern smart grids, causing significant economic losses and compromising grid reliability. This study introduces the SmartGuard Energy Intelligence System (SGEIS), an integrated artificial intelligence framework for electricity theft detection and intelligent energy monitoring. The proposed system combines supervised machine learning, deep learning-based time-series modeling, Non-Intrusive Load Monitoring (NILM), and graph-based learning to capture both temporal and spatial consumption patterns. A comprehensive data processing pipeline is developed, incorporating feature engineering, multi-scale temporal analysis, and rule-based anomaly labeling. Deep learning models, including Long Short-Term Memory (LSTM), Temporal Convolutional Networks (TCN), and Autoencoders, are employed to detect abnormal usage patterns. In parallel, ensemble learning methods such as Random Forest, Gradient Boosting, XGBoost, and LightGBM are utilized for classification. To model grid topology and spatial dependencies, Graph Neural Networks (GNNs) are applied to identify correlated anomalies across interconnected nodes. The NILM module enhances interpretability by disaggregating appliance-level consumption from aggregate signals. Experimental results demonstrate strong performance, with Gradient Boosting achieving a ROC-AUC of 0.894, while graph-based models attain over 96% accuracy in identifying high-risk nodes. The hybrid framework improves detection robustness by integrating temporal, statistical, and spatial intelligence. Overall, SGEIS provides a scalable and practical solution for electricity theft detection, offering high accuracy, improved interpretability, and strong potential for real-world smart grid deployment.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Explainable Model Routing for Agentic Workflows
arXiv:2604.03527v1 Announce Type: new
Abstract: Modern agentic workflows decompose complex tasks into specialized subtasks and route them to diverse models to minimize cost without sacrificing quality. However, current routing architectures focus exclusively on performance optimization, leaving underlying trade-offs between model capability and cost unrecorded. Without clear rationale, developers cannot distinguish between intelligent efficiency -- using specialized models for appropriate tasks -- and latent failures caused by budget-driven model selection. We present Topaz, a framework that introduces formal auditability to agentic routing. Topaz replaces silent model assignments with an inherently interpretable router that incorporates three components: (i) skill-based profiling that synthesizes performance across diverse benchmarks into granular capability profiles (ii) fully traceable routing algorithms that utilize budget-based and multi-objective optimization to produce clear traces of how skill-match scores were weighed against costs, and (iii) developer-facing explanations that translate these traces into natural language, allowing users to audit system logic and iteratively tune the cost-quality tradeoff. By making routing decisions interpretable, Topaz enables users to understand, trust, and meaningfully steer routed agentic systems.
Fonte: arXiv cs.AI
RL • Score 85
When Adaptive Rewards Hurt: Causal Probing and the Switching-Stability Dilemma in LLM-Guided LEO Satellite Scheduling
arXiv:2604.03562v1 Announce Type: new
Abstract: Adaptive reward design for deep reinforcement learning (DRL) in multi-beam LEO satellite scheduling is motivated by the intuition that regime-aware reward weights should outperform static ones. We systematically test this intuition and uncover a switching-stability dilemma: near-constant reward weights (342.1 Mbps) outperform carefully-tuned dynamic weights (103.3+/-96.8 Mbps) because PPO requires a quasistationary reward signal for value function convergence. Weight adaptation-regardless of quality-degrades performance by repeatedly restarting convergence. To understand why specific weights matter, we introduce a single-variable causal probing method that independently perturbs each reward term by +/-20% and measures PPO response after 50k steps. Probing reveals counterintuitive leverage: a +20% increase in the switching penalty yields +157 Mbps for polar handover and +130 Mbps for hot-cold regimes-findings inaccessible to human experts or trained MLPs without systematic probing. We evaluate four MDP architect variants (fixed, rule-based, learned MLP, finetuned LLM) across known and novel traffic regimes. The MLP achieves 357.9 Mbps on known regimes and 325.2 Mbps on novel regimes, while the fine-tuned LLM collapses to 45.3+/-43.0 Mbps due to weight oscillation rather than lack of domain knowledge-output consistency, not knowledge, is the binding constraint. Our findings provide an empirically-grounded roadmap for LLM-DRL integration in communication systems, identifying where LLMs add irreplaceable value (natural language intent understanding) versus where simpler methods suffice.
Fonte: arXiv cs.AI
RecSys • Score 85
Regime-Calibrated Demand Priors for Ride-Hailing Fleet Dispatch and Repositioning
arXiv:2604.03883v1 Announce Type: new
Abstract: Effective ride-hailing dispatch requires anticipating demand patterns that vary substantially across time-of-day, day-of-week, season, and special events. We propose a regime-calibrated approach that (i) segments historical trip data into demand regimes, (ii) matches the current operating period to the most similar historical analogues via a similarity ensemble combining Kolmogorov-Smirnov distance, Wasserstein-1 distance, feature distance, variance ratio, event pattern similarity, and temporal proximity, and (iii) uses the resulting calibrated demand prior to drive both an LP-based fleet repositioning policy and batch dispatch with Hungarian matching. In ablation, a distributional-only metric subset achieves the strongest mean-wait reduction, while the full ensemble is retained as a robustness-oriented default that preserves calendar and event context.
Evaluated on 5.2 million NYC TLC trips across 8 diverse scenarios (winter/summer, weekday/weekend/holiday, morning/evening/night) with 5 random seeds each, our method reduces mean rider wait times by 31.1% (bootstrap 95% CI: [26.5, 36.6]; Friedman chi-squared = 80.0, p = 4.25e-18; Cohen's d = 7.5-29.9). P95 wait drops 37.6% and the Gini coefficient of wait times improves from 0.441 to 0.409. The two contributions compose multiplicatively: calibration provides 16.9% reduction relative to the replay baseline; LP repositioning adds a further 15.5%. The approach requires no training, is deterministic and explainable, generalizes to Chicago (23.3% wait reduction using the NYC-built regime library without retraining), and is robust across fleet sizes (32-47% improvement for 0.5x-2.0x fleet scaling). Code is available at https://github.com/IndarKarhana/regime-calibrated-dispatch.
Fonte: arXiv cs.LG
Multimodal • Score 85
Solar-VLM: Multimodal Vision-Language Models for Augmented Solar Power Forecasting
arXiv:2604.04145v1 Announce Type: new
Abstract: Photovoltaic (PV) power forecasting plays a critical role in power system dispatch and market participation. Because PV generation is highly sensitive to weather conditions and cloud motion, accurate forecasting requires effective modeling of complex spatiotemporal dependencies across multiple information sources. Although recent studies have advanced AI-based forecasting methods, most fail to fuse temporal observations, satellite imagery, and textual weather information in a unified framework. This paper proposes Solar-VLM, a large-language-model-driven framework for multimodal PV power forecasting. First, modality-specific encoders are developed to extract complementary features from heterogeneous inputs. The time-series encoder adopts a patch-based design to capture temporal patterns from multivariate observations at each site. The visual encoder, built upon a Qwen-based vision backbone, extracts cloud-cover information from satellite images. The text encoder distills historical weather characteristics from textual descriptions. Second, to capture spatial dependencies across geographically distributed PV stations, a cross-site feature fusion mechanism is introduced. Specifically, a Graph Learner models inter-station correlations through a graph attention network constructed over a K-nearest-neighbor (KNN) graph, while a cross-site attention module further facilitates adaptive information exchange among sites. Finally, experiments conducted on data from eight PV stations in a northern province of China demonstrate the effectiveness of the proposed framework. Our proposed model is publicly available at https://github.com/rhp413/Solar-VLM.
Fonte: arXiv cs.AI
RL • Score 85
Provable Multi-Task Reinforcement Learning: A Representation Learning Framework with Low Rank Rewards
arXiv:2604.03891v1 Announce Type: new
Abstract: Multi-task representation learning (MTRL) is an approach that learns shared latent representations across related tasks, facilitating collaborative learning that improves the overall learning efficiency. This paper studies MTRL for multi-task reinforcement learning (RL), where multiple tasks have the same state-action space and transition probabilities, but different rewards. We consider T linear Markov Decision Processes (MDPs) where the reward functions and transition dynamics admit linear feature embeddings of dimension d. The relatedness among the tasks is captured by a low-rank structure on the reward matrices. Learning shared representations across multiple RL tasks is challenging due to the complex and policy-dependent nature of data that leads to a temporal progression of error. Our approach adopts a reward-free reinforcement learning framework to first learn a data-collection policy. This policy then informs an exploration strategy for estimating the unknown reward matrices. Importantly, the data collected under this well-designed policy enable accurate estimation, which ultimately supports the learning of an near-optimal policy. Unlike existing approaches that rely on restrictive assumptions such as Gaussian features, incoherence conditions, or access to optimal solutions, we propose a low-rank matrix estimation method that operates under more general feature distributions encountered in RL settings. Theoretical analysis establishes that accurate low-rank matrix recovery is achievable under these relaxed assumptions, and we characterize the relationship between representation error and sample complexity. Leveraging the learned representation, we construct near-optimal policies and prove a regret bound. Experimental results demonstrate that our method effectively learns robust shared representations and task dynamics from finite data.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Neural Global Optimization via Iterative Refinement from Noisy Samples
arXiv:2604.03614v1 Announce Type: new
Abstract: Global optimization of black-box functions from noisy samples is a fundamental challenge in machine learning and scientific computing. Traditional methods such as Bayesian Optimization often converge to local minima on multi-modal functions, while gradient-free methods require many function evaluations. We present a novel neural approach that learns to find global minima through iterative refinement. Our model takes noisy function samples and their fitted spline representation as input, then iteratively refines an initial guess toward the true global minimum. Trained on randomly generated functions with ground truth global minima obtained via exhaustive search, our method achieves a mean error of 8.05 percent on challenging multi-modal test functions, compared to 36.24 percent for the spline initialization, a 28.18 percent improvement. The model successfully finds global minima in 72 percent of test cases with error below 10 percent, demonstrating learned optimization principles rather than mere curve fitting. Our architecture combines encoding of multiple modalities including function values, derivatives, and spline coefficients with iterative position updates, enabling robust global optimization without requiring derivative information or multiple restarts.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Diagonal-Tiled Mixed-Precision Attention for Efficient Low-Bit MXFP Inference
arXiv:2604.03950v1 Announce Type: new
Abstract: Transformer-based large language models (LLMs) have demonstrated remarkable performance across a wide range of real-world tasks, but their inference cost remains prohibitively high due to the quadratic complexity of attention and the memory bandwidth limitations of high-precision operations. In this work, we present a low-bit mixed-precision attention kernel using the microscaling floating-point (MXFP) data format, utilizing the computing capability on next-generation GPU architectures. Our Diagonal-Tiled Mixed-Precision Attention (DMA) incorporates two kinds of low-bit computation at the tiling-level, and is a delicate fused kernel implemented using Triton, exploiting hardware-level parallelism and memory efficiency to enable fast and efficient inference without compromising model performance. Extensive empirical evaluations on NVIDIA B200 GPUs show that our kernel maintains generation quality with negligible degradation, and meanwhile achieves significant speedup by kernel fusion. We release our code at https://github.com/yifu-ding/MP-Sparse-Attn.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Zero-Shot Quantization via Weight-Space Arithmetic
arXiv:2604.03420v1 Announce Type: new
Abstract: We show that robustness to post-training quantization (PTQ) is a transferable direction in weight space. We call this direction the quantization vector: extracted from a donor task by simple weight-space arithmetic, it can be used to patch a receiver model and improve robustness to PTQ-induced noise by as much as 60%, without receiver-side quantization-aware training (QAT). Because the method requires no receiver training data, it provides a zero-shot, low-cost alternative to QAT for extremely low-bit deployment. We demonstrate this on Vision Transformer (ViT) models. More broadly, our results suggest that quantization robustness is not merely a byproduct of task-specific training, but a reusable feature of weight-space geometry that can be transferred rather than retrained.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
BWTA: Accurate and Efficient Binarized Transformer by Algorithm-Hardware Co-design
arXiv:2604.03957v1 Announce Type: new
Abstract: Ultra low-bit quantization brings substantial efficiency for Transformer-based models, but the accuracy degradation and limited GPU support hinder its wide usage. In this paper, we analyze zero-point distortion in binarization and propose a Binary Weights & Ternary Activations (BWTA) quantization scheme, which projects tiny values to zero and preserves the accuracy of extremely low-bit models. For training, we propose Smooth Multi-Stage Quantization, combining a Levelwise Degradation Strategy and a Magnitude-Alignment Projection Factor to enable stable and fast convergence. For inference, we develop a BWTA MatMul CUDA kernel with instruction-level parallel bit-packing and comprehensive binary/ternary MatMul implementations for both linear and attention operators, allowing seamless integration across Transformer architectures. Experiments show that BWTA approaches full-precision performance for BERT, with an average 3.5% drop on GLUE and less than 2% drop on five tasks, and achieves comparable perplexity and accuracy for LLMs. In efficiency, it delivers 16 to 24 times kernel-level speedup over FP16 on NVIDIA GPUs, and 216 to 330 tokens/s end-to-end prefill speedup with lower memory footprint on LLMs. As an algorithm-hardware co-design, BWTA demonstrates practical, low-latency ultra-low-bit inference without sacrificing model quality.
Fonte: arXiv cs.LG
RecSys • Score 85
Supervised Dimensionality Reduction Revisited: Why LDA on Frozen CNN Features Deserves a Second Look
arXiv:2604.03928v1 Announce Type: new
Abstract: Effective ride-hailing dispatch requires anticipating demand patterns that vary substantially across time-of-day, day-of-week, season, and special events. We propose a regime-calibrated approach that (i) segments historical trip data into demand regimes, (ii) matches the current operating period to the most similar historical analogues via a six-metric similarity ensemble (Kolmogorov-Smirnov, Wasserstein-1, feature distance, variance ratio, event pattern, temporal proximity), and (iii) uses the resulting calibrated demand prior to drive both an LP-based fleet repositioning policy and batch dispatch with Hungarian matching. In ablation, a distributional-only subset is strongest on mean wait, while the full ensemble is retained as a robustness-oriented default.
Evaluated on 5.2 million NYC TLC trips across 8 diverse scenarios (winter/summer, weekday/weekend/holiday, morning/evening/night) with 5 random seeds each, our method reduces mean rider wait times by 31.1% (bootstrap 95% CI: [26.5, 36.6]%; Friedman chi-sq = 80.0, p = 4.25e-18; Cohen's d = 7.5-29.9 across scenarios). The improvement extends to the tail: P95 wait drops 37.6% and the Gini coefficient of wait times improves from 0.441 to 0.409 (7.3% relative). The two contributions compose multiplicatively and are independently validated: calibration provides 16.9% reduction; LP repositioning adds a further 15.5%. The approach requires no training, is deterministic and explainable, generalizes to Chicago (23.3% wait reduction via NYC-built regime library), and is robust across fleet sizes (32-47% improvement for 0.5-2x fleet scaling). We provide comprehensive ablation studies, formal statistical tests, and routing-fidelity validation with OSRM.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Modeling and Controlling Deployment Reliability under Temporal Distribution Shift
arXiv:2604.02351v1 Announce Type: new
Abstract: Machine learning models deployed in non-stationary environments are exposed to temporal distribution shift, which can erode predictive reliability over time. While common mitigation strategies such as periodic retraining and recalibration aim to preserve performance, they typically focus on average metrics evaluated at isolated time points and do not explicitly model how reliability evolves during deployment.
We propose a deployment-centric framework that treats reliability as a dynamic state composed of discrimination and calibration. The trajectory of this state across sequential evaluation windows induces a measurable notion of volatility, allowing deployment adaptation to be formulated as a multi-objective control problem that balances reliability stability against cumulative intervention cost.
Within this framework, we define a family of state-dependent intervention policies and empirically characterize the resulting cost-volatility Pareto frontier. Experiments on a large-scale, temporally indexed credit-risk dataset (1.35M loans, 2007-2018) show that selective, drift-triggered interventions can achieve smoother reliability trajectories than continuous rolling retraining while substantially reducing operational cost.
These findings position deployment reliability under temporal shift as a controllable multi-objective system and highlight the role of policy design in shaping stability-cost trade-offs in high-stakes tabular applications.
Fonte: arXiv cs.LG
RecSys • Score 85
Learn then Decide: A Learning Approach for Designing Data Marketplaces
arXiv:2503.10773v2 Announce Type: replace
Abstract: As data marketplaces become increasingly central to the digital economy, it is crucial to design efficient pricing mechanisms that optimize revenue while ensuring fair and adaptive pricing. We introduce the Maximum Auction-to-Posted Price (MAPP) mechanism, a novel two-stage approach that first estimates the bidders' value distribution through auctions and then determines the optimal posted price based on the learned distribution. We establish that MAPP is individually rational and incentive-compatible, ensuring truthful bidding while balancing revenue maximization with minimal price discrimination. On the theoretical side, we establish a statistical viewpoint that recasts revenue optimization as a valuation density estimation problem: we show that revenue regret can be controlled by uniform error in estimating the valuation density. MAPP achieves a regret of $O_p(n^{-1}(\log n)^2)$ when incorporating historical bid data, where $n$ is the number of bids in the current round. For sequential dataset sales over $T$ rounds, we propose an online MAPP mechanism that dynamically adjusts pricing across datasets with varying value distributions. Our approach achieves no-regret learning, with the average cumulative regret converging at a rate of $O_p(T^{-1/2}(\log T)^2)$. We validate the effectiveness of MAPP through simulations and real-world data from the FCC AWS-3 spectrum auction.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
UI-Oceanus: Scaling GUI Agents with Synthetic Environmental Dynamics
arXiv:2604.02345v1 Announce Type: cross
Abstract: Scaling generalist GUI agents is hindered by the data scalability bottleneck of expensive human demonstrations and the "distillation ceiling" of synthetic teacher supervision. To transcend these limitations, we propose UI-Oceanus, a framework that shifts the learning focus from mimicking high-level trajectories to mastering interaction physics via ground-truth environmental feedback. Through a systematic investigation of self-supervised objectives, we identify that forward dynamics, defined as the generative prediction of future interface states, acts as the primary driver for scalability and significantly outweighs inverse inference. UI-Oceanus leverages this insight by converting low-cost autonomous exploration, which is verified directly by system execution, into high-density generative supervision to construct a robust internal world model. Experimental evaluations across a series of models demonstrate the decisive superiority of our approach: models utilizing Continual Pre-Training (CPT) on synthetic dynamics outperform non-CPT baselines with an average success rate improvement of 7% on offline benchmarks, which amplifies to a 16.8% gain in real-world online navigation. Furthermore, we observe that navigation performance scales with synthetic data volume. These results confirm that grounding agents in forward predictive modeling offers a superior pathway to scalable GUI automation with robust cross-domain adaptability and compositional generalization.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
JoyAI-LLM Flash: Advancing Mid-Scale LLMs with Token Efficiency
arXiv:2604.03044v1 Announce Type: new
Abstract: We introduce JoyAI-LLM Flash, an efficient Mixture-of-Experts (MoE) language model designed to redefine the trade-off between strong performance and token efficiency in the sub-50B parameter regime. JoyAI-LLM Flash is pretrained on a massive corpus of 20 trillion tokens and further optimized through a rigorous post-training pipeline, including supervised fine-tuning (SFT), Direct Preference Optimization (DPO), and large-scale reinforcement learning (RL) across diverse environments. To improve token efficiency, JoyAI-LLM Flash strategically balances \emph{thinking} and \emph{non-thinking} cognitive modes and introduces FiberPO, a novel RL algorithm inspired by fibration theory that decomposes trust-region maintenance into global and local components, providing unified multi-scale stability control for LLM policy optimization. To enhance architectural sparsity, the model comprises 48B total parameters while activating only 2.7B parameters per forward pass, achieving a substantially higher sparsity ratio than contemporary industry leading models of comparable scale. To further improve inference throughput, we adopt a joint training-inference co-design that incorporates dense Multi-Token Prediction (MTP) and Quantization-Aware Training (QAT). We release the checkpoints for both JoyAI-LLM-48B-A3B Base and its post-trained variants on Hugging Face to support the open-source community.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
Holos: A Web-Scale LLM-Based Multi-Agent System for the Agentic Web
arXiv:2604.02334v1 Announce Type: new
Abstract: As large language models (LLM)-driven agents transition from isolated task solvers to persistent digital entities, the emergence of the Agentic Web, an ecosystem where heterogeneous agents autonomously interact and co-evolve, marks a pivotal shift toward Artificial General Intelligence (AGI). However, LLM-based multi-agent systems (LaMAS) are hindered by open-world issues such as scaling friction, coordination breakdown, and value dissipation. To address these challenges, we introduce Holos, a web-scale LaMAS architected for long-term ecological persistence. Holos adopts a five-layer architecture, with core modules primarily featuring the Nuwa engine for high-efficiency agent generation and hosting, a market-driven Orchestrator for resilient coordination, and an endogenous value cycle to achieve incentive compatibility. By bridging the gap between micro-level collaboration and macro-scale emergence, Holos hopes to lay the foundation for the next generation of the self-organizing and continuously evolving Agentic Web. We have publicly released Holos (accessible at https://holosai.io), providing a resource for the community and a testbed for future research in large-scale agentic ecosystems.
Fonte: arXiv cs.AI
RL • Score 85
Coupled Control, Structured Memory, and Verifiable Action in Agentic AI (SCRAT -- Stochastic Control with Retrieval and Auditable Trajectories): A Comparative Perspective from Squirrel Locomotion and Scatter-Hoarding
arXiv:2604.03201v1 Announce Type: new
Abstract: Agentic AI is increasingly judged not by fluent output alone but by whether it can act, remember, and verify under partial observability, delay, and strategic observation. Existing research often studies these demands separately: robotics emphasizes control, retrieval systems emphasize memory, and alignment or assurance work emphasizes checking and oversight. This article argues that squirrel ecology offers a sharp comparative case because arboreal locomotion, scatter-hoarding, and audience-sensitive caching couple all three demands in one organism. We synthesize evidence from fox, eastern gray, and, in one field comparison, red squirrels, and impose an explicit inference ladder: empirical observation, minimal computational inference, and AI design conjecture. We introduce a minimal hierarchical partially observed control model with latent dynamics, structured episodic memory, observer-belief state, option-level actions, and delayed verifier signals. This motivates three hypotheses: (H1) fast local feedback plus predictive compensation improves robustness under hidden dynamics shifts; (H2) memory organized for future control improves delayed retrieval under cue conflict and load; and (H3) verifiers and observer models inside the action-memory loop reduce silent failure and information leakage while remaining vulnerable to misspecification. A downstream conjecture is that role-differentiated proposer/executor/checker/adversary systems may reduce correlated error under asymmetric information and verification burden. The contribution is a comparative perspective and benchmark agenda: a disciplined program of falsifiable claims about the coupling of control, memory, and verifiable action.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
Structure-Preserving Multi-View Embedding Using Gromov-Wasserstein Optimal Transport
arXiv:2604.02610v1 Announce Type: new
Abstract: Multi-view data analysis seeks to integrate multiple representations of the same samples in order to recover a coherent low-dimensional structure. Classical approaches often rely on feature concatenation or explicit alignment assumptions, which become restrictive under heterogeneous geometries or nonlinear distortions. In this work, we propose two geometry-aware multi-view embedding strategies grounded in Gromov-Wasserstein (GW) optimal transport. The first, termed Mean-GWMDS, aggregates view-specific relational information by averaging distance matrices and applying GW-based multidimensional scaling to obtain a representative embedding. The second strategy, referred to as Multi-GWMDS, adopts a selection-based paradigm in which multiple geometry-consistent candidate embeddings are generated via GW-based alignment and a representative embedding is selected. Experiments on synthetic manifolds and real-world datasets show that the proposed methods effectively preserve intrinsic relational structure across views. These results highlight GW-based approaches as a flexible and principled framework for multi-view representation learning.
Fonte: arXiv stat.ML
Theory/Optimization • Score 85
Lipschitz bounds for integral kernels
arXiv:2604.02887v1 Announce Type: new
Abstract: Feature maps associated with positive definite kernels play a central role in kernel methods and learning theory, where regularity properties such as Lipschitz continuity are closely related to robustness and stability guarantees. Despite their importance, explicit characterizations of the Lipschitz constant of kernel feature maps are available only in a limited number of cases. In this paper, we study the Lipschitz regularity of feature maps associated with integral kernels under differentiability assumptions. We first provide sufficient conditions ensuring Lipschitz continuity and derive explicit formulas for the corresponding Lipschitz constants. We then identify a condition under which the feature map fails to be Lipschitz continuous and apply these results to several important classes of kernels. For infinite width two-layer neural network with isotropic Gaussian weight distributions, we show that the Lipschitz constant of the associated kernel can be expressed as the supremum of a two-dimensional integral, leading to an explicit characterization for the Gaussian kernel and the ReLU random neural network kernel. We also study continuous and shift-invariant kernels such as Gaussian, Laplace, and Mat\'ern kernels, which admit an interpretation as neural network with cosine activation function. In this setting, we prove that the feature map is Lipschitz continuous if and only if the weight distribution has a finite second-order moment, and we then derive its Lipschitz constant. Finally, we raise an open question concerning the asymptotic behavior of the convergence of the Lipschitz constant in finite width neural networks. Numerical experiments are provided to support this behavior.
Fonte: arXiv stat.ML
Multimodal • Score 85
Token-Efficient Multimodal Reasoning via Image Prompt Packaging
arXiv:2604.02492v1 Announce Type: new
Abstract: Deploying large multimodal language models at scale is constrained by token-based inference costs, yet the cost-performance behavior of visual prompting strategies remains poorly characterized. We introduce Image Prompt Packaging (IPPg), a prompting paradigm that embeds structured text directly into images to reduce text token overhead, and benchmark it across five datasets, three frontier models (GPT-4.1, GPT-4o, Claude 3.5 Sonnet), and two task families (VQA and code generation). We derive a cost formulation decomposing savings by token type and show IPPg achieves 35.8--91.0\% inference cost reductions. Despite token compression of up to 96\%, accuracy remains competitive in many settings, though outcomes are highly model- and task-dependent: GPT-4.1 achieves simultaneous accuracy and cost gains on CoSQL, while Claude 3.5 incurs cost increases on several VQA benchmarks. Systematic error analysis yields a failure-mode taxonomy: spatial reasoning, non-English inputs, and character-sensitive operations are most vulnerable, while schema-structured tasks benefit most. A 125-configuration rendering ablation reveals accuracy shifts of 10--30 percentage points, establishing visual encoding choices as a first-class variable in multimodal system design.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
A Numerical Method for Coupling Parameterized Physics-Informed Neural Networks and FDM for Advanced Thermal-Hydraulic System Simulation
arXiv:2604.02663v1 Announce Type: new
Abstract: Severe accident analysis using system-level codes such as MELCOR is indispensable for nuclear safety assessment, yet the computational cost of repeated simulations poses a significant bottleneck for parametric studies and uncertainty quantification. Existing surrogate models accelerate these analyses but depend on large volumes of simulation data, while physics-informed neural networks (PINNs) enable data-free training but must be retrained for every change in problem parameters. This study addresses both limitations by developing the Parameterized PINNs coupled with FDM (P2F) method, a node-assigned hybrid framework for MELCOR's Control Volume Hydrodynamics/Flow Path (CVH/FP) module. In the P2F method, a parameterized Node-Assigned PINN (NA-PINN) accepts the water-level difference, initial velocity, and time as inputs, learning a solution manifold so that a single trained network serves as a data-free surrogate for the momentum conservation equation across all flow paths without retraining. This PINN is coupled with a finite difference method (FDM) solver that advances the mass conservation equation at each time step, ensuring exact discrete mass conservation while replacing the iterative nonlinear momentum solve with a single forward pass. Verification on a six-tank gravity-driven draining scenario yields a water level mean absolute error of $7.85 \times 10^{-5}$ m and a velocity mean absolute error of $3.21 \times 10^{-3}$ m/s under the nominal condition with $\Delta t = 1.0$ s. The framework maintains consistent accuracy across time steps ranging from 0.2 to 1.0 s and generalizes to five distinct initial conditions, all without retraining or simulation data. This work introduces a numerical coupling methodology for integrating parameterized PINNs with FDM within a nuclear thermal-hydraulic system code framework.
Fonte: arXiv cs.LG
Vision • Score 85
Generating Satellite Imagery Data for Wildfire Detection through Mask-Conditioned Generative AI
arXiv:2604.02479v1 Announce Type: new
Abstract: The scarcity of labeled satellite imagery remains a fundamental bottleneck for deep-learning (DL)-based wildfire monitoring systems. This paper investigates whether a diffusion-based foundation model for Earth Observation (EO), EarthSynth, can synthesize realistic post-wildfire Sentinel-2 RGB imagery conditioned on existing burn masks, without task-specific retraining.
Using burn masks derived from the CalFireSeg-50 dataset (Martin et al., 2025), we design and evaluate six controlled experimental configurations that systematically vary: (i) pipeline architecture (mask-only full generation vs. inpainting with pre-fire context), (ii) prompt engineering strategy (three hand-crafted prompts and a VLM-generated prompt via Qwen2-VL), and (iii) a region-wise color-matching post-processing step.
Quantitative assessment on 10 stratified test samples uses four complementary metrics: Burn IoU, burn-region color distance ({\Delta}C_burn), Darkness Contrast, and Spectral Plausibility. Results show that inpainting-based pipelines consistently outperform full-tile generation across all metrics, with the structured inpainting prompt achieving the best spatial alignment (Burn IoU = 0.456) and burn saliency (Darkness Contrast = 20.44), while color matching produces the lowest color distance ({\Delta}C_burn = 63.22) at the cost of reduced burn saliency.
VLM-assisted inpainting is competitive with hand-crafted prompts. These findings provide a foundation for incorporating generative data augmentation into wildfire detection pipelines.
Code and experiments are available at: https://www.kaggle.com/code/valeriamartinh/genai-all-runned
Fonte: arXiv cs.CV
Theory/Optimization • Score 85
WGFINNs: Weak formulation-based GENERIC formalism informed neural networks'
arXiv:2604.02601v1 Announce Type: new
Abstract: Data-driven discovery of governing equations from noisy observations remains a fundamental challenge in scientific machine learning. While GENERIC formalism informed neural networks (GFINNs) provide a principled framework that enforces the laws of thermodynamics by construction, their reliance on strong-form loss formulations makes them highly sensitive to measurement noise. To address this limitation, we propose weak formulation-based GENERIC formalism informed neural networks (WGFINNs), which integrate the weak formulation of dynamical systems with the structure-preserving architecture of GFINNs. WGFINNs significantly enhance robustness to noisy data while retaining exact satisfaction of GENERIC degeneracy and symmetry conditions. We further incorporate a state-wise weighted loss and a residual-based attention mechanism to mitigate scale imbalance across state variables. Theoretical analysis contrasts quantitative differences between the strong-form and the weak-form estimators. Mainly, the strong-form estimator diverges as the time step decreases in the presence of noise, while the weak-form estimator can be accurate even with noisy data if test functions satisfy certain conditions. Numerical experiments demonstrate that WGFINNs consistently outperform GFINNs at varying noise levels, achieving more accurate predictions and reliable recovery of physical quantities.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
TRACE: Traceroute-based Internet Route change Analysis with Ensemble Learning
arXiv:2604.02361v1 Announce Type: cross
Abstract: Detecting Internet routing instability is a critical yet challenging task, particularly when relying solely on endpoint active measurements. This study introduces TRACE, a MachineLearning (ML)pipeline designed to identify route changes using only traceroute latency data, thereby ensuring independence from control plane information. We propose a robust feature engineering strategy that captures temporal dynamics using rolling statistics and aggregated context patterns. The architecture leverages a stacked ensemble of Gradient Boosted Decision Trees refined by a hyperparameter-optimized meta-learner. By strictly calibrating decision thresholds to address the inherent class imbalance of rare routing events, TRACE achieves a superior F1-score performance, significantly outperforming traditional baseline models and demonstrating strong effective ness in detecting routing changes on the Internet.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
State estimations and noise identifications with intermittent corrupted observations via Bayesian variational inference
arXiv:2604.02738v1 Announce Type: new
Abstract: This paper focuses on the state estimation problem in distributed sensor networks, where intermittent packet dropouts, corrupted observations, and unknown noise covariances coexist. To tackle this challenge, we formulate the joint estimation of system states, noise parameters, and network reliability as a Bayesian variational inference problem, and propose a novel variational Bayesian adaptive Kalman filter (VB-AKF) to approximate the joint posterior probability densities of the latent parameters. Unlike existing AKF that separately handle missing data and measurement outliers, the proposed VB-AKF adopts a dual-mask generative model with two independent Bernoulli random variables, explicitly characterizing both observable communication losses and latent data authenticity. Additionally, the VB-AKF integrates multiple concurrent multiple observations into the adaptive filtering framework, which significantly enhances statistical identifiability. Comprehensive numerical experiments verify the effectiveness and asymptotic optimality of the proposed method, showing that both parameter identification and state estimation asymptotically converge to the theoretical optimal lower bound with the increase in the number of sensors.
Fonte: arXiv stat.ML
MLOps/Systems • Score 85
Characterizing WebGPU Dispatch Overhead for LLM Inference Across Four GPU Vendors, Three Backends, and Three Browsers
arXiv:2604.02344v1 Announce Type: new
Abstract: WebGPU's security-focused design imposes per-operation validation that compounds across the many small dispatches in neural network inference, yet the true cost of this overhead is poorly characterized. We present a systematic characterization of WebGPU dispatch overhead for LLM inference at batch size 1, spanning four GPU vendors (NVIDIA, AMD, Apple, Intel), two native implementations (Dawn, wgpu-native) and three browsers (Chrome, Safari, Firefox), and two model sizes (Qwen2.5-0.5B and 1.5B). Our primary contribution is a sequential-dispatch methodology that reveals naive single-operation benchmarks overestimate dispatch cost by ${\sim}20\times$. The true per-dispatch cost of WebGPU API overhead alone is 24-36 $\mu$s on Vulkan and 32-71 $\mu$s on Metal, while the total per-operation overhead including Python cost is ${\sim}95$~$\mu$s, which turns out to be a distinction critical for optimization. On Vulkan, kernel fusion improves throughput by 53%, while CUDA fusion provides no benefit, confirming that per-operation overhead is a primary differentiator. LLM inference was tested across three major operating systems (Linux, Windows, macOS). We built $\texttt{torch-webgpu}$, a PrivateUse1-based out-of-tree PyTorch backend and an FX-to-WebGPU compiler, which on our reference platform achieves 11--12% of CUDA performance. At dtype-matched float32, RTX PRO 2000 achieves 1.4$\times$ WebGPU's throughput despite ${\sim}6\times$ less compute than RTX 5090. For dispatch overhead, backend choice is the dominant factor, although implementation choice also matters substantially within a backend (2.2$\times$ for Metal). In terms of dispatch vs kernel compute efficiency, we conclude that at batch=1 with the current dispatch-heavy pipeline, per-operation overhead dominates regardless of kernel quality. All code, benchmarks, and raw data are open source.
Fonte: arXiv cs.LG
RL • Score 85
Dynamic Mask Enhanced Intelligent Multi-UAV Deployment for Urban Vehicular Networks
arXiv:2604.02358v1 Announce Type: cross
Abstract: Vehicular Ad Hoc Networks (VANETs) play a crucial role in realizing vehicle-road collaboration and intelligent transportation. However, urban VANETs often face challenges such as frequent link disconnections and subnet fragmentation, which hinder reliable connectivity. To address these issues, we dynamically deploy multiple Unmanned Aerial Vehicles (UAVs) as communication relays to enhance VANET. A novel Score based Dynamic Action Mask enhanced QMIX algorithm (Q-SDAM) is proposed for multi-UAV deployment, which maximizes vehicle connectivity while minimizing multi-UAV energy consumption. Specifically, we design a score-based dynamic action mask mechanism to guide UAV agents in exploring large action spaces, accelerate the learning process and enhance optimization performance. The practicality of Q-SDAM is validated using real-world datasets. We show that Q-SDAM improves connectivity by 18.2% while reducing energy consumption by 66.6% compared with existing algorithms.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
AdaHOP: Fast and Accurate Low-Precision Training via Outlier-Pattern-Aware Rotation
arXiv:2604.02525v1 Announce Type: new
Abstract: Low-precision training (LPT) commonly employs Hadamard transforms to suppress outliers and mitigate quantization error in large language models (LLMs). However, prior methods apply a fixed transform uniformly, despite substantial variation in outlier structures across tensors. Through the first systematic study of outlier patterns across weights, activations, and gradients of LLMs, we show that this strategy is fundamentally flawed: the effectiveness of Hadamard-based suppression depends on how the transform's smoothing direction aligns with the outlier structure of each operand -- a property that varies substantially across layers and computation paths. We characterize these patterns into three types: Row-wise, Column-wise, and None. Each pair requires a tailored transform direction or outlier handling strategy to minimize quantization error. Based on this insight, we propose AdaHOP (Adaptive Hadamard transform with Outlier-Pattern-aware strategy), which assigns each matrix multiplication its optimal strategy: Inner Hadamard Transform (IHT) where inner-dimension smoothing is effective, or IHT combined with selective Outlier Extraction (OE) -- routing dominant outliers to a high-precision path -- where it is not. Combined with hardware-aware Triton kernels, AdaHOP achieves BF16 training quality at MXFP4 precision while delivering up to 3.6X memory compression and 1.8X kernel acceleration} over BF16 full-precision training.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
SEDGE: Structural Extrapolated Data Generation
arXiv:2604.02482v1 Announce Type: new
Abstract: This paper proposes a framework for Structural Extrapolated Data GEneration (SEDGE) based on suitable assumptions on the underlying data generating process. We provide conditions under which data satisfying new specifications can be generated reliably, together with the approximate identifiability of the distribution of such data under certain ``conservative" assumptions. On the algorithmic side, we develop practical methods to achieve extrapolated data generation, based on the structure-informed optimization strategy or diffusion posterior sampling, respectively. We verify the extrapolation performance on synthetic data and also consider extrapolated image generation as a real-world scenario to illustrate the validity of the proposed framework.
Fonte: arXiv cs.LG
Applications • Score 85
LumiVideo: An Intelligent Agentic System for Video Color Grading
arXiv:2604.02409v1 Announce Type: new
Abstract: Video color grading is a critical post-production process that transforms flat, log-encoded raw footage into emotionally resonant cinematic visuals. Existing automated methods act as static, black-box executors that directly output edited pixels, lacking both interpretability and the iterative control required by professionals. We introduce LumiVideo, an agentic system that mimics the cognitive workflow of professional colorists through four stages: Perception, Reasoning, Execution, and Reflection. Given only raw log video, LumiVideo autonomously produces a cinematic base grade by analyzing the scene's physical lighting and semantic content. Its Reasoning engine synergizes an LLM's internalized cinematic knowledge with a Retrieval-Augmented Generation (RAG) framework via a Tree of Thoughts (ToT) search to navigate the non-linear color parameter space. Rather than generating pixels, the system compiles the deduced parameters into industry-standard ASC-CDL configurations and a globally consistent 3D LUT, analytically guaranteeing temporal consistency. An optional Reflection loop then allows creators to refine the result via natural language feedback. We further introduce LumiGrade, the first log-encoded video benchmark for evaluating automated grading. Experiments show that LumiVideo approaches human expert quality in fully automatic mode while enabling precise iterative control when directed.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
Convolutional Surrogate for 3D Discrete Fracture-Matrix Tensor Upscaling
arXiv:2604.02335v1 Announce Type: new
Abstract: Modeling groundwater flow in three-dimensional fractured crystalline media requires accounting for strong spatial heterogeneity induced by fractures. Fine-scale discrete fracture-matrix (DFM) simulations can capture this complexity but are computationally expensive, especially when repeated evaluations are needed. To address this, we aim to employ a multilevel Monte Carlo (MLMC) framework in which numerical homogenization is used to upscale sub-resolution fracture effects when transitioning between accuracy levels.
To reduce the cost of conventional 3D numerical homogenization, we develop a surrogate model that predicts the equivalent hydraulic conductivity tensor Keq from a voxelized 3D domain representing tensor-valued random fields of matrix and fracture conductivities. Fracture size, orientation, and aperture are sampled from distributions informed by natural observations.
The surrogate architecture combines a 3D convolutional neural network with feed-forward layers, enabling it to capture both local spatial features and global interactions. Three surrogates are trained on data generated by DFM simulations, each corresponding to a different fracture-to-matrix conductivity contrast. Performance is evaluated across a wide range of fracture network parameters and matrix-field correlation lengths.
The trained models achieve high accuracy, with normalized root-mean-square errors below 0.22 across most test cases. Practical applicability is demonstrated by comparing numerically homogenized conductivities with surrogate predictions in two macro-scale problems: computing equivalent conductivity tensors and predicting outflow from a constrained 3D domain. In both cases, surrogate-based upscaling preserves accuracy while substantially reducing computational cost, achieving speedups exceeding 100x when inference is performed on a GPU.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Reinforcement Learning-based Knowledge Distillation with LLM-as-a-Judge
arXiv:2604.02621v1 Announce Type: new
Abstract: Reinforcement Learning (RL) has been shown to substantially improve the reasoning capability of small and large language models (LLMs), but existing approaches typically rely on verifiable rewards, hence ground truth labels. We propose an RL framework that uses rewards from an LLM that acts as a judge evaluating model outputs over large amounts of unlabeled data, enabling label-free knowledge distillation and replacing the need of ground truth supervision. Notably, the judge operates with a single-token output, making reward computation efficient. When combined with verifiable rewards, our approach yields substantial performance gains across math reasoning benchmarks. These results suggest that LLM-based evaluators can produce effective training signals for RL fine-tuning.
Fonte: arXiv cs.CL
RL • Score 85
Mitigating Data Scarcity in Spaceflight Applications for Offline Reinforcement Learning Using Physics-Informed Deep Generative Models
arXiv:2604.02438v1 Announce Type: new
Abstract: The deployment of reinforcement learning (RL)-based controllers on physical systems is often limited by poor generalization to real-world scenarios, known as the simulation-to-reality (sim-to-real) gap. This gap is particularly challenging in spaceflight, where real-world training data are scarce due to high cost and limited planetary exploration data. Traditional approaches, such as system identification and synthetic data generation, depend on sufficient data and often fail due to modeling assumptions or lack of physics-based constraints. We propose addressing this data scarcity by introducing physics-based learning bias in a generative model. Specifically, we develop the Mutual Information-based Split Variational Autoencoder (MI-VAE), a physics-informed VAE that learns differences between observed system trajectories and those predicted by physics-based models. The latent space of the MI-VAE enables generation of synthetic datasets that respect physical constraints. We evaluate MI-VAE on a planetary lander problem, focusing on limited real-world data and offline RL training. Results show that augmenting datasets with MI-VAE samples significantly improves downstream RL performance, outperforming standard VAEs in statistical fidelity, sample diversity, and policy success rate. This work demonstrates a scalable strategy for enhancing autonomous controller robustness in complex, data-constrained environments.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Fast NF4 Dequantization Kernels for Large Language Model Inference
arXiv:2604.02556v1 Announce Type: new
Abstract: Large language models (LLMs) have grown beyond the memory capacity of single GPU devices, necessitating quantization techniques for practical deployment. While NF4 (4-bit NormalFloat) quantization enables 4$\times$ memory reduction, inference on current NVIDIA GPUs (e.g., Ampere A100) requires expensive dequantization back to FP16 format, creating a critical performance bottleneck. This paper presents a lightweight shared memory optimization that addresses this gap through principled memory hierarchy exploitation while maintaining full ecosystem compatibility. We compare our technique against the open-source BitsAndBytes implementation, achieving 2.0--2.2$\times$ kernel speedup across three models (Gemma 27B, Qwen3 32B, and Llama3.3 70B) and up to 1.54$\times$ end-to-end improvement by leveraging the 12--15$\times$ latency advantage of shared memory over global memory access. Our optimization reduces instruction counts through simplified indexing logic while using only 64 bytes of shared memory per thread block, demonstrating that lightweight optimizations can deliver substantial performance gains with minimal engineering effort. This work provides a plug-and-play solution for the HuggingFace ecosystem that democratizes access to advanced models on existing GPU infrastructure.
Fonte: arXiv cs.LG
Vision • Score 85
WSVD: Weighted Low-Rank Approximation for Fast and Efficient Execution of Low-Precision Vision-Language Models
arXiv:2604.02570v1 Announce Type: new
Abstract: Singular Value Decomposition (SVD) has become an important technique for reducing the computational burden of Vision Language Models (VLMs), which play a central role in tasks such as image captioning and visual question answering. Although multiple prior works have proposed efficient SVD variants to enable low-rank operations, we find that in practice it remains difficult to achieve substantial latency reduction during model execution. To address this limitation, we introduce a new computational pattern and apply SVD at a finer granularity, enabling real and measurable improvements in execution latency. Furthermore, recognizing that weight elements differ in their relative importance, we adaptively allocate relative importance to each element during SVD process to better preserve accuracy, then extend this framework with quantization applied to both weights and activations, resulting in a highly efficient VLM. Collectively, we introduce~\textit{Weighted SVD} (WSVD), which outperforms other approaches by achieving over $1.8\times$ decoding speedup while preserving accuracy. We open source our code at: \href{https://github.com/SAI-Lab-NYU/WSVD}{\texttt{https://github.com/SAI-Lab-NYU/WSVD}
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Parser-Oriented Structural Refinement for a Stable Layout Interface in Document Parsing
arXiv:2604.02692v1 Announce Type: new
Abstract: Accurate document parsing requires both robust content recognition and a stable parser interface. In explicit Document Layout Analysis (DLA) pipelines, downstream parsers do not consume the full detector output. Instead, they operate on a retained and serialized set of layout instances. However, on dense pages with overlapping regions and ambiguous boundaries, unstable layout hypotheses can make the retained instance set inconsistent with its parser input order, leading to severe downstream parsing errors. To address this issue, we introduce a lightweight structural refinement stage between a DETR-style detector and the parser to stabilize the parser interface. Treating raw detector outputs as a compact hypothesis pool, the proposed module performs set-level reasoning over query features, semantic cues, box geometry, and visual evidence. From a shared refined structural state, it jointly determines instance retention, refines box localization, and predicts parser input order before handoff. We further introduce retention-oriented supervision and a difficulty-aware ordering objective to better align the retained instance set and its order with the final parser input, especially on structurally complex pages. Extensive experiments on public benchmarks show that our method consistently improves page-level layout quality. When integrated into a standard end-to-end parsing pipeline, the stabilized parser interface also substantially reduces sequence mismatch, achieving a Reading Order Edit of 0.024 on OmniDocBench.
Fonte: arXiv cs.CV
Vision • Score 90
ExploreVLA: Dense World Modeling and Exploration for End-to-End Autonomous Driving
arXiv:2604.02714v1 Announce Type: new
Abstract: End-to-end autonomous driving models based on Vision-Language-Action (VLA) architectures have shown promising results by learning driving policies through behavior cloning on expert demonstrations. However, imitation learning inherently limits the model to replicating observed behaviors without exploring diverse driving strategies, leaving it brittle in novel or out-of-distribution scenarios. Reinforcement learning (RL) offers a natural remedy by enabling policy exploration beyond the expert distribution. Yet VLA models, typically trained on offline datasets, lack directly observable state transitions, necessitating a learned world model to anticipate action consequences. In this work, we propose a unified understanding-and-generation framework that leverages world modeling to simultaneously enable meaningful exploration and provide dense supervision. Specifically, we augment trajectory prediction with future RGB and depth image generation as dense world modeling objectives, requiring the model to learn fine-grained visual and geometric representations that substantially enrich the planning backbone. Beyond serving as a supervisory signal, the world model further acts as a source of intrinsic reward for policy exploration: its image prediction uncertainty naturally measures a trajectory's novelty relative to the training distribution, where high uncertainty indicates out-of-distribution scenarios that, if safe, represent valuable learning opportunities. We incorporate this exploration signal into a safety-gated reward and optimize the policy via Group Relative Policy Optimization (GRPO). Experiments on the NAVSIM and nuScenes benchmarks demonstrate the effectiveness of our approach, achieving a state-of-the-art PDMS score of 93.7 and an EPDMS of 88.8 on NAVSIM. The code and demo will be publicly available at https://zihaosheng.github.io/ExploreVLA/.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Haiku to Opus in Just 10 bits: LLMs Unlock Massive Compression Gains
arXiv:2604.02343v1 Announce Type: cross
Abstract: We study the compression of LLM-generated text across lossless and lossy regimes, characterizing a compression-compute frontier where more compression is possible at the cost of more compute. For lossless compression, domain-adapted LoRA adapters can improve LLM-based arithmetic coding by 2x over compression with the base LLM alone. For lossy compression, prompting a model for a succinct rewrite then applying arithmetic coding can achieve compression ratios of approximately 0.03, a 2x improvement over compressing the original response.
We further introduce Question-Asking compression (QA), an interactive lossy protocol inspired by the game 'Twenty Questions'. A small model iteratively refines its response by asking yes/no questions to a stronger model, transferring exactly one bit per answer. On 8 benchmarks spanning math, science, and code, 10 binary questions recover 23% to 72% of the capability gap between a small and large model on standard benchmarks and 7% to 38% on harder benchmarks, achieving compression ratios of 0.0006 to 0.004. This is over 100x smaller than prior LLM-based compression (Deletang et al., 2024), suggesting that interactive protocols can transfer knowledge far more efficiently than transmitting full responses.
Fonte: arXiv cs.AI
Privacy/Security/Fairness • Score 85
Homophily-aware Supervised Contrastive Counterfactual Augmented Fair Graph Neural Network
arXiv:2604.02342v1 Announce Type: new
Abstract: In recent years, Graph Neural Networks (GNNs) have achieved remarkable success in tasks such as node classification, link prediction, and graph representation learning. However, they remain susceptible to biases that can arise not only from node attributes but also from the graph structure itself. Addressing fairness in GNNs has therefore emerged as a critical research challenge. In this work, we propose a novel model for training fairness-aware GNNs by improving the counterfactual augmented fair graph neural network framework (CAF). Specifically, our approach introduces a two-phase training strategy: in the first phase, we edit the graph to increase homophily ratio with respect to class labels while reducing homophily ratio with respect to sensitive attribute labels; in the second phase, we integrate a modified supervised contrastive loss and environmental loss into the optimization process, enabling the model to jointly improve predictive performance and fairness. Experiments on five real-world datasets demonstrate that our model outperforms CAF and several state-of-the-art graph-based learning methods in both classification accuracy and fairness metrics.
Fonte: arXiv cs.LG
Multimodal • Score 85
LiME: Lightweight Mixture of Experts for Efficient Multimodal Multi-task Learning
arXiv:2604.02338v1 Announce Type: new
Abstract: MoE-PEFT methods combine Mixture of Experts with parameter-efficient fine-tuning for multi-task adaptation, but require separate adapters per expert causing trainable parameters to scale linearly with expert count and limiting applicability to adapter-based architectures. We propose LiME (Lightweight Mixture of Experts), which achieves expert specialization through lightweight modulation rather than adapter replication. Instead of separate adapters, LiME uses a single shared PEFT module and modulates its output with lightweight expert vectors, reducing expert parameters while generalizing to any PEFT method. Notably, LiME introduces zero-parameter routing by leveraging existing frozen and adapted representations eliminating learned router parameters typically required per layer. Theoretically, we prove that (i) more experts preserve more task-relevant information and (ii) modulation approximates full expert-specific PEFT with bounded error. LiME further incorporates n-gram windowed routing and adaptive expert selection (Auto Top-K) based on routing confidence. Experiments on MMT-47, a multimodal multi-task benchmark with 47 tasks spanning text, image, and video, demonstrate that LiME achieves competitive or superior performance while using up to 4x fewer trainable parameters and up to 29% faster training compared to corresponding MoE-PEFT baselines.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
AXELRAM: Quantize Once, Never Dequantize
arXiv:2604.02638v1 Announce Type: new
Abstract: We propose AXELRAM, a smart SRAM macro architecture that computes attention scores directly from quantized KV cache indices without dequantization. The key enabler is a design-time fixed codebook: orthogonal-transform-based quantization concentrates each coordinate's distribution to N(0,1/d), so the optimal quantizer depends only on dimension d and bit-width b, not on input data. The asymmetric path design -- transform on write, table-lookup on read with no inverse transform -- reduces per-query multiplications by 102.4x (a mathematical identity).
Through multi-seed evaluation (10 seeds x 3 models), we discover that sign pattern sensitivity causes catastrophic PPL spikes (Delta > 50) on certain models (Qwen2.5-3B), while others (LLaMA-3.1-8B) are fully stable. This phenomenon extends SpinQuant's observation of rotation variance in weight quantization to the KV cache domain, where the effect is qualitatively more severe. We trace the root cause to layer-wise norm heterogeneity and propose a gradient-free sign pattern selection (200 candidates, 8 calibration samples, one-time) that eliminates catastrophic spikes with zero additional hardware cost. All source code is available at https://github.com/Axelidea/AXELRAM.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Low-Rank Compression of Pretrained Models via Randomized Subspace Iteration
arXiv:2604.02659v1 Announce Type: new
Abstract: The massive scale of pretrained models has made efficient compression essential for practical deployment. Low-rank decomposition based on the singular value decomposition (SVD) provides a principled approach for model reduction, but its exact computation is expensive for large weight matrices. Randomized alternatives such as randomized SVD (RSVD) improve efficiency, yet they can suffer from poor approximation quality when the singular value spectrum decays slowly, a regime commonly observed in modern pretrained models. In this work, we address this limitation from both theoretical and empirical perspectives. First, we establish a connection between low-rank approximation error and predictive performance by analyzing softmax perturbations, showing that deviations in class probabilities are controlled by the spectral error of the compressed weights. Second, we demonstrate that RSVD is inadequate, and we propose randomized subspace iteration (RSI) as a more effective alternative. By incorporating multiple power iterations, RSI improves spectral separation and provides a controllable mechanism for enhancing approximation quality. We evaluate our approach on both convolutional networks and transformer-based architectures. Our results show that RSI achieves near-optimal approximation quality while outperforming RSVD in predictive accuracy under aggressive compression, enabling efficient model compression.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
FTimeXer: Frequency-aware Time-series Transformer with Exogenous variables for Robust Carbon Footprint Forecasting
arXiv:2604.02347v1 Announce Type: new
Abstract: Accurate and up-to-date forecasting of the power grid's carbon footprint is crucial for effective product carbon footprint (PCF) accounting and informed decarbonization decisions. However, the carbon intensity of the grid exhibits high non-stationarity, and existing methods often struggle to effectively leverage periodic and oscillatory patterns. Furthermore, these methods tend to perform poorly when confronted with irregular exogenous inputs, such as missing data or misalignment. To tackle these challenges, we propose FTimeXer, a frequency-aware time-series Transformer designed with a robust training scheme that accommodates exogenous factors. FTimeXer features an Fast Fourier Transform (FFT)-driven frequency branch combined with gated time-frequency fusion, allowing it to capture multi-scale periodicity effectively. It also employs stochastic exogenous masking in conjunction with consistency regularization, which helps reduce spurious correlations and enhance stability. Experiments conducted on three real-world datasets show consistent improvements over strong baselines. As a result, these enhancements lead to more reliable forecasts of grid carbon factors, which are essential for effective PCF accounting and informed decision-making regarding decarbonization.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Automatic Textbook Formalization
arXiv:2604.03071v1 Announce Type: new
Abstract: We present a case study where an automatic AI system formalizes a textbook with more than 500 pages of graduate-level algebraic combinatorics to Lean. The resulting formalization represents a new milestone in textbook formalization scale and proficiency, moving from early results in undergraduate topology and restructuring of existing library content to a full standalone formalization of a graduate textbook. The formalization comprises 130K lines of code and 5900 Lean declarations and was conducted within one week by a total of 30K Claude 4.5 Opus agents collaborating in parallel on a shared code base via version control, simultaneously setting a record in multi-agent software engineering with usable results. The inference cost matches or undercuts what we estimate as the salaries required for a team of human experts, and we expect there is still the potential for large efficiencies to be made without the need for better models. We make our code, the resulting Lean code base and a side-by-side blueprint website available open-source.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
InfoSeeker: A Scalable Hierarchical Parallel Agent Framework for Web Information Seeking
arXiv:2604.02971v1 Announce Type: new
Abstract: Recent agentic search systems have made substantial progress by emphasising deep, multi-step reasoning. However, this focus often overlooks the challenges of wide-scale information synthesis, where agents must aggregate large volumes of heterogeneous evidence across many sources. As a result, most existing large language model agent systems face severe limitations in data-intensive settings, including context saturation, cascading error propagation, and high end-to-end latency. To address these challenges, we present \framework, a hierarchical framework based on principle of near-decomposability, containing a strategic \textit{Host}, multiple \textit{Managers} and parallel \textit{Workers}. By leveraging aggregation and reflection mechanisms at the Manager layer, our framework enforces strict context isolation to prevent saturation and error propagation. Simultaneously, the parallelism in worker layer accelerates the speed of overall task execution, mitigating the significant latency. Our evaluation on two complementary benchmarks demonstrates both efficiency ($ 3-5 \times$ speed-up) and effectiveness, achieving a $8.4\%$ success rate on WideSearch-en and $52.9\%$ accuracy on BrowseComp-zh. The code is released at https://github.com/agent-on-the-fly/InfoSeeker
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
CharTool: Tool-Integrated Visual Reasoning for Chart Understanding
arXiv:2604.02794v1 Announce Type: new
Abstract: Charts are ubiquitous in scientific and financial literature for presenting structured data. However, chart reasoning remains challenging for multimodal large language models (MLLMs) due to the lack of high-quality training data, as well as the need for fine-grained visual grounding and precise numerical computation. To address these challenges, we first propose DuoChart, a scalable dual-source data pipeline that combines synthesized charts with real-world charts to construct diverse, high-quality chart training data. We then introduce CharTool, which equips MLLMs with external tools, including image cropping for localized visual perception and code-based computation for accurate numerical reasoning. Through agentic reinforcement learning on DuoChart, CharTool learns tool-integrated reasoning grounded in chart content. Extensive experiments on six chart benchmarks show that our method consistently improves over strong MLLM baselines across model scales. Notably, CharTool-7B outperforms the base model by **+8.0%** on CharXiv (Reasoning) and **+9.78%** on ChartQAPro, while achieving competitive performance with substantially larger or proprietary models. Moreover, CharTool demonstrates positive generalization to out-of-domain visual math reasoning benchmarks.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
VBGS-SLAM: Variational Bayesian Gaussian Splatting Simultaneous Localization and Mapping
arXiv:2604.02696v1 Announce Type: new
Abstract: 3D Gaussian Splatting (3DGS) has shown promising results for 3D scene modeling using mixtures of Gaussians, yet its existing simultaneous localization and mapping (SLAM) variants typically rely on direct, deterministic pose optimization against the splat map, making them sensitive to initialization and susceptible to catastrophic forgetting as map evolves. We propose Variational Bayesian Gaussian Splatting SLAM (VBGS-SLAM), a novel framework that couples the splat map refinement and camera pose tracking in a generative probabilistic form. By leveraging conjugate properties of multivariate Gaussians and variational inference, our method admits efficient closed-form updates and explicitly maintains posterior uncertainty over both poses and scene parameters. This uncertainty-aware method mitigates drift and enhances robustness in challenging conditions, while preserving the efficiency and rendering quality of existing 3DGS. Our experiments demonstrate superior tracking performance and robustness in long sequence prediction, alongside efficient, high-quality novel view synthesis across diverse synthetic and real-world scenes.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Revealing the Learning Dynamics of Long-Context Continual Pre-training
arXiv:2604.02650v1 Announce Type: new
Abstract: Existing studies on Long-Context Continual Pre-training (LCCP) mainly focus on small-scale models and limited data regimes (tens of billions of tokens). We argue that directly migrating these small-scale settings to industrial-grade models risks insufficient adaptation and premature training termination. Furthermore, current evaluation methods rely heavily on downstream benchmarks (e.g., Needle-in-a-Haystack), which often fail to reflect the intrinsic convergence state and can lead to "deceptive saturation". In this paper, we present the first systematic investigation of LCCP learning dynamics using the industrial-grade Hunyuan-A13B (80B total parameters), tracking its evolution across a 200B-token training trajectory. Specifically, we propose a hierarchical framework to analyze LCCP dynamics across behavioral (supervised fine-tuning probing), probabilistic (perplexity), and mechanistic (attention patterns) levels. Our findings reveal: (1) Necessity of Massive Data Scaling: Training regimes of dozens of billions of tokens are insufficient for industrial-grade LLMs' LCCP (e.g., Hunyuan-A13B reaches saturation after training over 150B tokens). (2) Deceptive Saturation vs. Intrinsic Saturation: Traditional NIAH scores report "fake saturation" early, while our PPL-based analysis reveals continuous intrinsic improvements and correlates more strongly with downstream performance. (3) Mechanistic Monitoring for Training Stability: Retrieval heads act as efficient, low-resource training monitors, as their evolving attention scores reliably track LCCP progress and exhibit high correlation with SFT results. This work provides a comprehensive monitoring framework, evaluation system, and mechanistic interpretation for the LCCP of industrial-grade LLM.
Fonte: arXiv cs.CL
Privacy/Security/Fairness • Score 85
Communication-Efficient Distributed Learning with Differential Privacy
arXiv:2604.02558v1 Announce Type: new
Abstract: We address nonconvex learning problems over undirected networks. In particular, we focus on the challenge of designing an algorithm that is both communication-efficient and that guarantees the privacy of the agents' data. The first goal is achieved through a local training approach, which reduces communication frequency. The second goal is achieved by perturbing gradients during local training, specifically through gradient clipping and additive noise. We prove that the resulting algorithm converges to a stationary point of the problem within a bounded distance. Additionally, we provide theoretical privacy guarantees within a differential privacy framework that ensure agents' training data cannot be inferred from the trained model shared over the network. We show the algorithm's superior performance on a classification task under the same privacy budget, compared with state-of-the-art methods.
Fonte: arXiv cs.LG
Privacy/Security/Fairness • Score 85
Unlocking Multi-Site Clinical Data: A Federated Approach to Privacy-First Child Autism Behavior Analysis
arXiv:2604.02616v1 Announce Type: new
Abstract: Automated recognition of autistic behaviors in children is essential for early intervention and objective clinical assessment. However, the development of robust models is severely hindered by strict privacy regulations (e.g., HIPAA) and the sensitive nature of pediatric data, which prevents the centralized aggregation of clinical datasets. Furthermore, individual clinical sites often suffer from data scarcity, making it difficult to learn generalized behavior patterns or tailor models to site-specific patient distributions. To address these challenges, we observe that Federated Learning (FL) can decouple model training from raw data access, enabling multi-site collaboration while maintaining strict data residency. In this paper, we present the first study exploring Federated Learning for pose-based child autism behavior recognition. Our framework employs a two-layer privacy protection mechanism: utilizing human skeletal abstraction to remove identifiable visual information from the raw RGB videos and FL to ensure sensitive pose data remains within the clinic. This approach leverages distributed clinical data to learn generalized representations while providing the flexibility for site-specific personalization. Experimental results on the MMASD benchmark demonstrate that our framework achieves high recognition accuracy, outperforming traditional federated baselines and providing a robust, privacy-first solution for multi-site clinical analysis.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
AIVV: Neuro-Symbolic LLM Agent-Integrated Verification and Validation for Trustworthy Autonomous Systems
arXiv:2604.02478v1 Announce Type: new
Abstract: Deep learning models excel at detecting anomaly patterns in normal data. However, they do not provide a direct solution for anomaly classification and scalability across diverse control systems, frequently failing to distinguish genuine faults from nuisance faults caused by noise or the control system's large transient response. Consequently, because algorithmic fault validation remains unscalable, full Verification and Validation (V\&V) operations are still managed by Human-in-the-Loop (HITL) analysis, resulting in an unsustainable manual workload. To automate this essential oversight, we propose Agent-Integrated Verification and Validation (AIVV), a hybrid framework that deploys Large Language Models (LLMs) as a deliberative outer loop. Because rigorous system verification strictly depends on accurate validation, AIVV escalates mathematically flagged anomalies to a role-specialized LLM council. The council agents perform collaborative validation by semantically validating nuisance and true failures based on natural-language (NL) requirements to secure a high-fidelity system-verification baseline. Building on this foundation, the council then performs system verification by assessing post-fault responses against NL operational tolerances, ultimately generating actionable V\&V artifacts, such as gain-tuning proposals. Experiments on a time-series simulator for Unmanned Underwater Vehicles (UUVs) demonstrate that AIVV successfully digitizes the HITL V\&V process, overcoming the limitations of rule-based fault classification and offering a scalable blueprint for LLM-mediated oversight in time-series data domains.
Fonte: arXiv cs.AI
Multimodal • Score 85
Efficient3D: A Unified Framework for Adaptive and Debiased Token Reduction in 3D MLLMs
arXiv:2604.02689v1 Announce Type: new
Abstract: Recent advances in Multimodal Large Language Models (MLLMs) have expanded reasoning capabilities into 3D domains, enabling fine-grained spatial understanding. However, the substantial size of 3D MLLMs and the high dimensionality of input features introduce considerable inference overhead, which limits practical deployment on resource constrained platforms. To overcome this limitation, this paper presents Efficient3D, a unified framework for visual token pruning that accelerates 3D MLLMs while maintaining competitive accuracy. The proposed framework introduces a Debiased Visual Token Importance Estimator (DVTIE) module, which considers the influence of shallow initial layers during attention aggregation, thereby producing more reliable importance predictions for visual tokens. In addition, an Adaptive Token Rebalancing (ATR) strategy is developed to dynamically adjust pruning strength based on scene complexity, preserving semantic completeness and maintaining balanced attention across layers. Together, they enable context-aware token reduction that maintains essential semantics with lower computation. Comprehensive experiments conducted on five representative 3D vision and language benchmarks, including ScanRefer, Multi3DRefer, Scan2Cap, ScanQA, and SQA3D, demonstrate that Efficient3D achieves superior performance compared with unpruned baselines, with a +2.57% CIDEr improvement on the Scan2Cap dataset. Therefore, Efficient3D provides a scalable and effective solution for efficient inference in 3D MLLMs. The code is released at: https://github.com/sol924/Efficient3D
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
QAPruner: Quantization-Aware Vision Token Pruning for Multimodal Large Language Models
arXiv:2604.02816v1 Announce Type: new
Abstract: Multimodal Large Language Models (MLLMs) have shown strong reasoning ability, but their high computational and memory costs hinder deployment in resource-constrained settings. While Post-Training Quantization (PTQ) and vision token pruning are standard compression techniques, they are usually treated as independent optimizations. In this paper, we show that these two techniques are strongly coupled: naively applying semantic-based token pruning to PTQ-optimized MLLMs can discard activation outliers that are important for numerical stability and thus worsen quantization errors in low-bit regimes (\textit{e.g.}, W4A4). To address this issue, we propose a quantization-aware vision token pruning framework. Our method introduces a lightweight hybrid sensitivity metric that combines simulated group-wise quantization error with outlier intensity. By combining this metric with standard semantic relevance scores, the method retains tokens that are both semantically informative and robust to quantization. Experiments on standard LLaVA architectures show that our method consistently outperforms naive integration baselines. At an aggressive pruning ratio that retains only 12.5\% of visual tokens, our framework improves accuracy by 2.24\% over the baseline and even surpasses dense quantization without pruning. To the best of our knowledge, this is the first method that explicitly co-optimizes vision token pruning and PTQ for accurate low-bit MLLM inference.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
A Survey on AI for 6G: Challenges and Opportunities
arXiv:2604.02370v1 Announce Type: cross
Abstract: As wireless communication evolves, each generation of networks brings new technologies that change how we connect and interact. Artificial Intelligence (AI) is becoming crucial in shaping the future of sixth-generation (6G) networks. By combining AI and Machine Learning (ML), 6G aims to offer high data rates, low latency, and extensive connectivity for applications including smart cities, autonomous systems, holographic telepresence, and the tactile internet. This paper provides a detailed overview of the role of AI in supporting 6G networks. It focuses on key technologies like deep learning, reinforcement learning, federated learning, and explainable AI. It also looks at how AI integrates with essential network functions and discusses challenges related to scalability, security, and energy efficiency, along with new solutions. Additionally, this work highlights perspectives that connect AI-driven analytics to 6G service domains like Ultra-Reliable Low-Latency Communication (URLLC), Enhanced Mobile Broadband (eMBB), Massive Machine-Type Communication (mMTC), and Integrated Sensing and Communication (ISAC). It addresses concerns about standardization, ethics, and sustainability. By summarizing recent research trends and identifying future directions, this survey offers a valuable reference for researchers and practitioners at the intersection of AI and next-generation wireless communication.
Fonte: arXiv cs.AI
Vision • Score 85
MOMO: Mars Orbital Model Foundation Model for Mars Orbital Applications
arXiv:2604.02719v1 Announce Type: new
Abstract: We introduce MOMO, the first multi-sensor foundation model for Mars remote sensing. MOMO uses model merge to integrate representations learned independently from three key Martian sensors (HiRISE, CTX, and THEMIS), spanning resolutions from 0.25 m/pixel to 100 m/pixel. Central to our method is our novel Equal Validation Loss (EVL) strategy, which aligns checkpoints across sensors based on validation loss similarity before fusion via task arithmetic. This ensures models are merged at compatible convergence stages, leading to improved stability and generalization. We train MOMO on a large-scale, high-quality corpus of $\sim 12$ million samples curated from Mars orbital data and evaluate it on 9 downstream tasks from Mars-Bench. MOMO achieves better overall performance compared to ImageNet pre-trained, earth observation foundation model, sensor-specific pre-training, and fully-supervised baselines. Particularly on segmentation tasks, MOMO shows consistent and significant performance improvement. Our results demonstrate that model merging through an optimal checkpoint selection strategy provides an effective approach for building foundation models for multi-resolution data. The model weights, pretraining code, pretraining data, and evaluation code are available at: https://github.com/kerner-lab/MOMO.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
Communication-free Sampling and 4D Hybrid Parallelism for Scalable Mini-batch GNN Training
arXiv:2604.02651v1 Announce Type: new
Abstract: Graph neural networks (GNNs) are widely used for learning on graph datasets derived from various real-world scenarios. Learning from extremely large graphs requires distributed training, and mini-batching with sampling is a popular approach for parallelizing GNN training. Existing distributed mini-batch approaches have significant performance bottlenecks due to expensive sampling methods and limited scaling when using data parallelism. In this work, we present ScaleGNN, a 4D parallel framework for scalable mini-batch GNN training that combines communication-free distributed sampling, 3D parallel matrix multiplication (PMM), and data parallelism. ScaleGNN introduces a uniform vertex sampling algorithm, enabling each process (GPU device) to construct its local mini-batch, i.e., subgraph partitions without any inter-process communication. 3D PMM enables scaling mini-batch training to much larger GPU counts than vanilla data parallelism with significantly lower communication overheads. We also present additional optimizations to overlap sampling with training, reduce communication overhead by sending data in lower precision, kernel fusion, and communication-computation overlap. We evaluate ScaleGNN on five graph datasets and demonstrate strong scaling up to 2048 GPUs on Perlmutter, 2048 GCDs on Frontier, and 1024 GPUs on Tuolumne. On Perlmutter, ScaleGNN achieves 3.5x end-to-end training speedup over the SOTA baseline on ogbn-products.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Ambig-IaC: Multi-level Disambiguation for Interactive Cloud Infrastructure-as-Code Synthesis
arXiv:2604.02382v1 Announce Type: cross
Abstract: The scale and complexity of modern cloud infrastructure have made Infrastructure-as-Code (IaC) essential for managing deployments. While large Language models (LLMs) are increasingly being used to generate IaC configurations from natural language, user requests are often underspecified. Unlike traditional code generation, IaC configurations cannot be executed cheaply or iteratively repaired, forcing the LLMs into an almost one-shot regime. We observe that ambiguity in IaC exhibits a tractable compositional structure: configurations decompose into three hierarchical axes (resources, topology, attributes) where higher-level decisions constrain lower-level ones. We propose a training-free, disagreement-driven framework that generates diverse candidate specifications, identifies structural disagreements across these axes, ranks them by informativeness, and produces targeted clarification questions that progressively narrow the configuration space. We introduce \textsc{Ambig-IaC}, a benchmark of 300 validated IaC tasks with ambiguous prompts, and an evaluation framework based on graph edit distance and embedding similarity. Our method outperforms the strongest baseline, achieving relative improvements of +18.4\% and +25.4\% on structure and attribute evaluations, respectively.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
An Initial Exploration of Contrastive Prompt Tuning to Generate Energy-Efficient Code
arXiv:2604.02352v1 Announce Type: cross
Abstract: Although LLMs are capable of generating functionally correct code, they also tend to produce less energy-efficient code in comparison to human-written solutions. As these inefficiencies lead to higher computational overhead, they are in direct conflict with Green Software Development (GSD) efforts, which aim to reduce the energy consumption of code. To support these efforts, this study aims to investigate whether and how LLMs can be optimized to promote the generation of energy-efficient code. To this end, we employ Contrastive Prompt Tuning (CPT). CPT combines Contrastive Learning techniques, which help the model to distinguish between efficient and inefficient code, and Prompt Tuning, a Parameter-Efficient Fine Tuning (PEFT) approach that requires only a fraction of the cost of traditional fine tuning. This study evaluates CPT on Python, Java and C++ coding problems across three different models to provide a comprehensive evaluation. The method achieves consistent improvements in code accuracy for two models but efficiency gains vary by model, language and task complexity, indicating that improvements are not uniformly reliable.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
Swift-SVD: Theoretical Optimality Meets Practical Efficiency in Low-Rank LLM Compression
arXiv:2604.01609v1 Announce Type: new
Abstract: The deployment of Large Language Models is constrained by the memory and bandwidth demands of static weights and dynamic Key-Value cache. SVD-based compression provides a hardware-friendly solution to reduce these costs. However, existing methods suffer from two key limitations: some are suboptimal in reconstruction error, while others are theoretically optimal but practically inefficient. In this paper, we propose Swift-SVD, an activation-aware, closed-form compression framework that simultaneously guarantees theoretical optimum, practical efficiency and numerical stability. Swift-SVD incrementally aggregates covariance of output activations given a batch of inputs and performs a single eigenvalue decomposition after aggregation, enabling training-free, fast, and optimal layer-wise low-rank approximation. We employ effective rank to analyze local layer-wise compressibility and design a dynamic rank allocation strategy that jointly accounts for local reconstruction loss and end-to-end layer importance. Extensive experiments across six LLMs and eight datasets demonstrate that Swift-SVD outperforms state-of-the-art baselines, achieving optimal compression accuracy while delivering 3-70X speedups in end-to-end compression time. Our code will be released upon acceptance.
Fonte: arXiv cs.CL
RL • Score 85
DEFT: Distribution-guided Efficient Fine-Tuning for Human Alignment
arXiv:2604.01787v1 Announce Type: new
Abstract: Reinforcement Learning from Human Feedback (RLHF), using algorithms like Proximal Policy Optimization (PPO), aligns Large Language Models (LLMs) with human values but is costly and unstable. Alternatives have been proposed to replace PPO or integrate Supervised Fine-Tuning (SFT) and contrastive learning for direct fine-tuning and value alignment. However, these methods still require voluminous data to learn preferences and may weaken the generalization ability of LLMs. To further enhance alignment efficiency and performance while mitigating the loss of generalization ability, this paper introduces Distribution-guided Efficient Fine-Tuning (DEFT), an efficient alignment framework incorporating data filtering and distributional guidance by calculating the differential distribution reward based on the output distribution of language model and the discrepancy distribution of preference data. A small yet high-quality subset is filtered from the raw data using a differential distribution reward, which is then incorporated into existing alignment methods to guide the model's output distribution. Experimental results demonstrate that the methods enhanced by DEFT outperform the original methods in both alignment capability and generalization ability, with significantly reduced training time.
Fonte: arXiv cs.CL
MLOps/Systems • Score 85
Prime Once, then Reprogram Locally: An Efficient Alternative to Black-Box Service Model Adaptation
arXiv:2604.01474v1 Announce Type: new
Abstract: Adapting closed-box service models (i.e., APIs) for target tasks typically relies on reprogramming via Zeroth-Order Optimization (ZOO). However, this standard strategy is known for extensive, costly API calls and often suffers from slow, unstable optimization. Furthermore, we observe that this paradigm faces new challenges with modern APIs (e.g., GPT-4o). These models can be less sensitive to the input perturbations ZOO relies on, thereby hindering performance gains. To address these limitations, we propose an Alternative efficient Reprogramming approach for Service models (AReS). Instead of direct, continuous closed-box optimization, AReS initiates a single-pass interaction with the service API to prime an amenable local pre-trained encoder. This priming stage trains only a lightweight layer on top of the local encoder, making it highly receptive to the subsequent glass-box (white-box) reprogramming stage performed directly on the local model. Consequently, all subsequent adaptation and inference rely solely on this local proxy, eliminating all further API costs. Experiments demonstrate AReS's effectiveness where prior ZOO-based methods struggle: on GPT-4o, AReS achieves a +27.8% gain over the zero-shot baseline, a task where ZOO-based methods provide little to no improvement. Broadly, across ten diverse datasets, AReS outperforms state-of-the-art methods (+2.5% for VLMs, +15.6% for standard VMs) while reducing API calls by over 99.99%. AReS thus provides a robust and practical solution for adapting modern closed-box models.
Fonte: arXiv cs.CV
RL • Score 85
Soft MPCritic: Amortized Model Predictive Value Iteration
arXiv:2604.01477v1 Announce Type: new
Abstract: Reinforcement learning (RL) and model predictive control (MPC) offer complementary strengths, yet combining them at scale remains computationally challenging. We propose soft MPCritic, an RL-MPC framework that learns in (soft) value space while using sample-based planning for both online control and value target generation. soft MPCritic instantiates MPC through model predictive path integral control (MPPI) and trains a terminal Q-function with fitted value iteration, aligning the learned value function with the planner and implicitly extending the effective planning horizon. We introduce an amortized warm-start strategy that recycles planned open-loop action sequences from online observations when computing batched MPPI-based value targets. This makes soft MPCritic computationally practical, while preserving solution quality. soft MPCritic plans in a scenario-based fashion with an ensemble of dynamic models trained for next-step prediction accuracy. Together, these ingredients enable soft MPCritic to learn effectively through robust, short-horizon planning on classic and complex control tasks. These results establish soft MPCritic as a practical and scalable blueprint for synthesizing MPC policies in settings where policy extraction and direct, long-horizon planning may fail.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
F3DGS: Federated 3D Gaussian Splatting for Decentralized Multi-Agent World Modeling
arXiv:2604.01605v1 Announce Type: new
Abstract: We present F3DGS, a federated 3D Gaussian Splatting framework for decentralized multi-agent 3D reconstruction. Existing 3DGS pipelines assume centralized access to all observations, which limits their applicability in distributed robotic settings where agents operate independently, and centralized data aggregation may be restricted. Directly extending centralized training to multi-agent systems introduces communication overhead and geometric inconsistency. F3DGS first constructs a shared geometric scaffold by registering locally merged LiDAR point clouds from multiple clients to initialize a global 3DGS model. During federated optimization, Gaussian positions are fixed to preserve geometric alignment, while each client updates only appearance-related attributes, including covariance, opacity, and spherical harmonic coefficients. The server aggregates these updates using visibility-aware aggregation, weighting each client's contribution by how frequently it observed each Gaussian, resolving the partial-observability challenge inherent to multi-agent exploration. To evaluate decentralized reconstruction, we collect a multi-sequence indoor dataset with synchronized LiDAR, RGB, and IMU measurements. Experiments show that F3DGS achieves reconstruction quality comparable to centralized training while enabling distributed optimization across agents. The dataset, development kit, and source code will be publicly released.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
NEMESIS: Noise-suppressed Efficient MAE with Enhanced Superpatch Integration Strategy
arXiv:2604.01612v1 Announce Type: new
Abstract: Volumetric CT imaging is essential for clinical diagnosis, yet annotating 3D volumes is expensive and time-consuming, motivating self-supervised learning (SSL) from unlabeled data. However, applying SSL to 3D CT remains challenging due to the high memory cost of full-volume transformers and the anisotropic spatial structure of CT data, which is not well captured by conventional masking strategies. We propose NEMESIS, a masked autoencoder (MAE) framework that operates on local 128x128x128 superpatches, enabling memory-efficient training while preserving anatomical detail. NEMESIS introduces three key components: (i) noise-enhanced reconstruction as a pretext task, (ii) Masked Anatomical Transformer Blocks (MATB) that perform dual-masking through parallel plane-wise and axis-wise token removal, and (iii) NEMESIS Tokens (NT) for cross-scale context aggregation. On the BTCV multi-organ classification benchmark, NEMESIS with a frozen backbone and a linear classifier achieves a mean AUROC of 0.9633, surpassing fully fine-tuned SuPreM (0.9493) and VoCo (0.9387). Under a low-label regime with only 10% of available annotations, it retains an AUROC of 0.9075, demonstrating strong label efficiency. Furthermore, the superpatch-based design reduces computational cost to 31.0 GFLOPs per forward pass, compared to 985.8 GFLOPs for the full-volume baseline, providing a scalable and robust foundation for 3D medical imaging.
Fonte: arXiv cs.CV
MLOps/Systems • Score 85
GPA: Learning GUI Process Automation from Demonstrations
arXiv:2604.01676v1 Announce Type: new
Abstract: GUI Process Automation (GPA) is a lightweight but general vision-based Robotic Process Automation (RPA), which enables fast and stable process replay with only a single demo. Addressing the fragility of traditional RPA and the non-deterministic risks of current vision language model-based GUI agents, GPA introduces three core benefits: (1) Robustness via Sequential Monte Carlo-based localization to handle rescaling and detection uncertainty; (2) Deterministic and Reliability safeguarded by readiness calibration; and (3) Privacy through fast, fully local execution. This approach delivers the adaptability, robustness, and security required for enterprise workflows. It can also be used as an MCP/CLI tool by other agents with coding capabilities so that the agent only reasons and orchestrates while GPA handles the GUI execution. We conducted a pilot experiment to compare GPA with Gemini 3 Pro (with CUA tools) and found that GPA achieves higher success rate with 10 times faster execution speed in finishing long-horizon GUI tasks.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Sparse Spectral LoRA: Routed Experts for Medical VLMs
arXiv:2604.01310v1 Announce Type: new
Abstract: Large vision-language models (VLMs) excel on general benchmarks but often lack robustness in medical imaging, where heterogeneous supervision induces cross-dataset interference and sensitivity to data regime (i.e., how the supervisory signals are mixed). In realistic clinical workflows, data and tasks arrive sequentially, so naive continual training further leads to catastrophic forgetting. To address these challenges, we propose MedQwen, a parameter-efficient medical VLM that couples a spectrally routed Mixture-of-Experts (MoE) with a theoretically grounded scaling rule that aligns low-rank updates with a full-rank, fully fine-tuned MoE, without changing the base architecture. Concretely, we initialize each expert from non-overlapping singular value decomposition (SVD) segments of the pretrained weight and introduce a residual compensation and scaling scheme to enable stable expert specialization and consistent routing under distribution shift. Across 23 medical datasets covering visual question answering, report generation, radiology classification, and hallucination mitigation, MedQwen achieves strong, reliable performance: it approaches full fine-tuning on zero-shot classification with 339$\times$ fewer trainable parameters, and reduces sequential forgetting to $\sim$5\% where strong baselines degrade by $>$20-50\%.
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Machine Learning for Network Attacks Classification and Statistical Evaluation of Adversarial Learning Methodologies for Synthetic Data Generation
arXiv:2603.17717v2 Announce Type: replace-cross
Abstract: Supervised detection of network attacks has always been a critical part of network intrusion detection systems (NIDS). Nowadays, in a pivotal time for artificial intelligence (AI), with even more sophisticated attacks that utilize advanced techniques, such as generative artificial intelligence (GenAI) and reinforcement learning, it has become a vital component if we wish to protect our personal data, which are scattered across the web. In this paper, we address two tasks, in the first unified multi-modal NIDS dataset, which incorporates flow-level data, packet payload information and temporal contextual features, from the reprocessed CIC-IDS-2017, CIC-IoT-2023, UNSW-NB15 and CIC-DDoS-2019, with the same feature space. In the first task we use machine learning (ML) algorithms, with stratified cross validation, in order to prevent network attacks, with stability and reliability. In the second task we use adversarial learning algorithms to generate synthetic data, compare them with the real ones and evaluate their fidelity, utility and privacy using the SDV framework, f-divergences, distinguishability and non-parametric statistical tests. The findings provide stable ML models for intrusion detection and generative models with high fidelity and utility, by combining the Synthetic Data Vault framework, the TRTS and TSTR tests, with non-parametric statistical tests and f-divergence measures.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Expert-Choice Routing Enables Adaptive Computation in Diffusion Language Models
arXiv:2604.01622v1 Announce Type: new
Abstract: Diffusion language models (DLMs) enable parallel, non-autoregressive text generation, yet existing DLM mixture-of-experts (MoE) models inherit token-choice (TC) routing from autoregressive systems, leading to load imbalance and rigid computation allocation. We show that expert-choice (EC) routing is a better fit for DLMs: it provides deterministic load balancing by design, yielding higher throughput and faster convergence than TC. Building on the property that EC capacity is externally controllable, we introduce timestep-dependent expert capacity, which varies expert allocation according to the denoising step. We find that allocating more capacity to low-mask-ratio steps consistently achieves the best performance under matched FLOPs, and provide a mechanistic explanation: tokens in low-mask-ratio contexts exhibit an order-of-magnitude higher learning efficiency, so concentrating compute on these steps yields the largest marginal return. Finally, we show that existing pretrained TC DLMs can be retrofitted to EC by replacing only the router, achieving faster convergence and improved accuracy across diverse downstream tasks. Together, these results establish EC routing as a superior paradigm for DLM MoE models and demonstrate that computation in DLMs can be treated as an adaptive policy rather than a fixed architectural constant. Code is available at https://github.com/zhangshuibai/EC-DLM.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Variational LSTM with Augmented Inputs: Nonlinear Response History Metamodeling with Aleatoric and Epistemic Uncertainty
arXiv:2604.01587v1 Announce Type: new
Abstract: Uncertainty propagation in high-dimensional nonlinear dynamic structural systems is pivotal in state-of-the-art performance-based design and risk assessment, where uncertainties from both excitations and structures, i.e., the aleatoric uncertainty, must be considered. This poses a significant challenge due to heavy computational demands. Machine learning techniques are thus introduced as metamodels to alleviate this burden. However, the "black box" nature of Machine learning models underscores the necessity of avoiding overly confident predictions, particularly when data and training efforts are insufficient. This creates a need, in addition to considering the aleatoric uncertainty, of estimating the uncertainty related to the prediction confidence, i.e., epistemic uncertainty, for machine learning-based metamodels. We developed a probabilistic metamodeling technique based on a variational long short-term memory (LSTM) with augmented inputs to simultaneously capture aleatoric and epistemic uncertainties. Key random system parameters are treated as augmented inputs alongside excitation series carrying record-to-record variability to capture the full range of aleatoric uncertainty. Meanwhile, epistemic uncertainty is effectively approximated via the Monte Carlo dropout scheme. Unlike computationally expensive full Bayesian approaches, this method incurs negligible additional training costs while enabling nearly cost-free uncertainty simulation. The proposed technique is demonstrated through multiple case studies involving stochastic seismic or wind excitations. Results show that the calibrated metamodels accurately reproduce nonlinear response time histories and provide confidence bounds indicating the associated epistemic uncertainty.
Fonte: arXiv cs.LG
RL • Score 85
Learning with Incomplete Context: Linear Contextual Bandits with Pretrained Imputation
arXiv:2510.09908v3 Announce Type: replace
Abstract: The rise of large-scale pretrained models has made it feasible to generate predictive or synthetic features at low cost, raising the question of how to incorporate such surrogate predictions into downstream decision-making. We study this problem in the setting of online linear contextual bandits, where contexts may be complex, nonstationary, and only partially observed. In addition to bandit data, we assume access to an auxiliary dataset containing fully observed contexts--common in practice since such data are collected without adaptive interventions. We propose PULSE-UCB, an algorithm that leverages pretrained models trained on the auxiliary data to impute missing features during online decision-making. We establish regret guarantees that decompose into a standard bandit term plus an additional component reflecting pretrained model quality. In the i.i.d. context case with H\"older-smooth missing features, PULSE-UCB achieves near-optimal performance, supported by matching lower bounds. Our results quantify how uncertainty in predicted contexts affects decision quality and how much historical data is needed to improve downstream learning.
Fonte: arXiv stat.ML
MLOps/Systems • Score 85
Towards Intrinsically Calibrated Uncertainty Quantification in Industrial Data-Driven Models via Diffusion Sampler
arXiv:2604.01870v1 Announce Type: new
Abstract: In modern process industries, data-driven models are important tools for real-time monitoring when key performance indicators are difficult to measure directly. While accurate predictions are essential, reliable uncertainty quantification (UQ) is equally critical for safety, reliability, and decision-making, but remains a major challenge in current data-driven approaches. In this work, we introduce a diffusion-based posterior sampling framework that inherently produces well-calibrated predictive uncertainty via faithful posterior sampling, eliminating the need for post-hoc calibration. In extensive evaluations on synthetic distributions, the Raman-based phenylacetic acid soft sensor benchmark, and a real ammonia synthesis case study, our method achieves practical improvements over existing UQ techniques in both uncertainty calibration and predictive accuracy. These results highlight diffusion samplers as a principled and scalable paradigm for advancing uncertainty-aware modeling in industrial applications.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
ZEUS: Accelerating Diffusion Models with Only Second-Order Predictor
arXiv:2604.01552v1 Announce Type: new
Abstract: Denoising generative models deliver high-fidelity generation but remain bottlenecked by inference latency due to the many iterative denoiser calls required during sampling. Training-free acceleration methods reduce latency by either sparsifying the model architecture or shortening the sampling trajectory. Current training-free acceleration methods are more complex than necessary: higher-order predictors amplify error under aggressive speedups, and architectural modifications hinder deployment. Beyond 2x acceleration, step skipping creates structural scarcity -- at most one fresh evaluation per local window -- leaving the computed output and its backward difference as the only causally grounded information. Based on this, we propose ZEUS, an acceleration method that predicts reduced denoiser evaluations using a second-order predictor, and stabilizes aggressive consecutive skipping with an interleaved scheme that avoids back-to-back extrapolations. ZEUS adds essentially zero overhead, no feature caches, and no architectural modifications, and it is compatible with different backbones, prediction objectives, and solver choices. Across image and video generation, ZEUS consistently improves the speed-fidelity performance over recent training-free baselines, achieving up to 3.2x end-to-end speedup while maintaining perceptual quality. Our code is available at: https://github.com/Ting-Justin-Jiang/ZEUS.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
CuTeGen: An LLM-Based Agentic Framework for Generation and Optimization of High-Performance GPU Kernels using CuTe
arXiv:2604.01489v1 Announce Type: new
Abstract: High-performance GPU kernels are critical to modern machine learning systems, yet developing efficient implementations remains a challenging, expert-driven process due to the tight coupling between algorithmic structure, memory hierarchy usage, and hardware-specific optimizations. Recent work has explored using large language models (LLMs) to generate GPU kernels automatically, but generated implementations often struggle to maintain correctness and achieve competitive performance across iterative refinements. We present CuTeGen, an agentic framework for automated generation and optimization of GPU kernels that treats kernel development as a structured generate--test--refine workflow. Unlike approaches that rely on one-shot generation or large-scale search over candidate implementations, CuTeGen focuses on progressive refinement of a single evolving kernel through execution-based validation, structured debugging, and staged optimization. A key design choice is to generate kernels using the CuTe abstraction layer, which exposes performance-critical structures such as tiling and data movement while providing a more stable representation for iterative modification. To guide performance improvement, CuTeGen incorporates workload-aware optimization prompts and delayed integration of profiling feedback. Experimental results on matrix multiplication and activation workloads demonstrate that the framework produces functionally correct kernels and achieves competitive performance relative to optimized library implementations.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
BTS-rPPG: Orthogonal Butterfly Temporal Shifting for Remote Photoplethysmography
arXiv:2604.01679v1 Announce Type: new
Abstract: Remote photoplethysmography (rPPG) enables contactless physiological sensing from facial videos by analyzing subtle appearance variations induced by blood circulation. However, modeling the temporal dynamics of these signals remains challenging, as many deep learning methods rely on temporal shifting or convolutional operators that aggregate information primarily from neighboring frames, resulting in predominantly local temporal modeling and limited temporal receptive fields. To address this limitation, we propose BTS-rPPG, a temporal modeling framework based on Orthogonal Butterfly Temporal Shifting (BTS). Inspired by the butterfly communication pattern in the Fast Fourier Transform (FFT), BTS establishes structured frame interactions via an XOR-based butterfly pairing schedule, progressively expanding the temporal receptive field and enabling efficient propagation of information across distant frames. Furthermore, we introduce an orthogonal feature transfer mechanism (OFT) that filters the source feature with respect to the target context before temporal shifting, retaining only the orthogonal component for cross-frame transmission. This reduces redundant feature propagation and encourages complementary temporal interaction. Extensive experiments on multiple benchmark datasets demonstrate that BTS-rPPG improves long-range temporal modeling of physiological dynamics and consistently outperforms existing temporal modeling strategies for rPPG estimation.
Fonte: arXiv cs.CV
Vision • Score 85
Automatic Image-Level Morphological Trait Annotation for Organismal Images
arXiv:2604.01619v1 Announce Type: new
Abstract: Morphological traits are physical characteristics of biological organisms that provide vital clues on how organisms interact with their environment. Yet extracting these traits remains a slow, expert-driven process, limiting their use in large-scale ecological studies. A major bottleneck is the absence of high-quality datasets linking biological images to trait-level annotations. In this work, we demonstrate that sparse autoencoders trained on foundation-model features yield monosemantic, spatially grounded neurons that consistently activate on meaningful morphological parts. Leveraging this property, we introduce a trait annotation pipeline that localizes salient regions and uses vision-language prompting to generate interpretable trait descriptions. Using this approach, we construct Bioscan-Traits, a dataset of 80K trait annotations spanning 19K insect images from BIOSCAN-5M. Human evaluation confirms the biological plausibility of the generated morphological descriptions. We assess design sensitivity through a comprehensive ablation study, systematically varying key design choices and measuring their impact on the quality of the resulting trait descriptions. By annotating traits with a modular pipeline rather than prohibitively expensive manual efforts, we offer a scalable way to inject biologically meaningful supervision into foundation models, enable large-scale morphological analyses, and bridge the gap between ecological relevance and machine-learning practicality.
Fonte: arXiv cs.CV
Vision • Score 85
HOT: Harmonic-Constrained Optimal Transport for Remote Photoplethysmography Domain Adaptation
arXiv:2604.01675v1 Announce Type: new
Abstract: Remote photoplethysmography (rPPG) enables non-contact physiological measurement from facial videos; however, its practical deployment is often hindered by substantial performance degradation under domain shift. While recent deep learning-based rPPG methods have achieved strong performance on individual datasets, they frequently overfit to appearance-related factors, such as illumination, camera characteristics, and color response, that vary significantly across domains. To address this limitation, we introduce frequency domain adaptation (FDA) as a principled strategy for modeling appearance variation in rPPG. By transferring low-frequency spectral components that encode domain-dependent appearance characteristics, FDA encourages rPPG models to learn invariance to appearance variations while retaining cardiac-induced signals. To further support physiologically consistent alignment under such appearance variation, we propose Harmonic-Constrained Optimal Transport (HOT), which leverages the harmonic property of cardiac signals to guide alignment between original and FDA-transferred representations. Extensive cross-dataset experiments demonstrate that the proposed FDA and HOT framework effectively enhances the robustness and generalization of rPPG models across diverse datasets.
Fonte: arXiv cs.CV
Theory/Optimization • Score 85
Sven: Singular Value Descent as a Computationally Efficient Natural Gradient Method
arXiv:2604.01279v1 Announce Type: new
Abstract: We introduce Sven (Singular Value dEsceNt), a new optimization algorithm for neural networks that exploits the natural decomposition of loss functions into a sum over individual data points, rather than reducing the full loss to a single scalar before computing a parameter update. Sven treats each data point's residual as a separate condition to be satisfied simultaneously, using the Moore-Penrose pseudoinverse of the loss Jacobian to find the minimum-norm parameter update that best satisfies all conditions at once. In practice, this pseudoinverse is approximated via a truncated singular value decomposition, retaining only the $k$ most significant directions and incurring a computational overhead of only a factor of $k$ relative to stochastic gradient descent. This is in comparison to traditional natural gradient methods, which scale as the square of the number of parameters. We show that Sven can be understood as a natural gradient method generalized to the over-parametrized regime, recovering natural gradient descent in the under-parametrized limit. On regression tasks, Sven significantly outperforms standard first-order methods including Adam, converging faster and to a lower final loss, while remaining competitive with LBFGS at a fraction of the wall-time cost. We discuss the primary challenge to scaling, namely memory overhead, and propose mitigation strategies. Beyond standard machine learning benchmarks, we anticipate that Sven will find natural application in scientific computing settings where custom loss functions decompose into several conditions.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
GradPower: Powering Gradients for Faster Language Model Pre-Training
arXiv:2505.24275v2 Announce Type: replace-cross
Abstract: We propose GradPower, a lightweight gradient-transformation technique for accelerating language model pre-training. Given a gradient vector $g=(g_i)_i$, GradPower first applies the elementwise sign-power transformation: $\varphi_p(g)=({\rm sign}(g_i)|g_i|^p)_{i}$ for a fixed $p>0$, and then feeds the transformed gradient into a base optimizer. Notably, GradPower requires only a single-line code change and no modifications to the base optimizer's internal logic, including the hyperparameters. When applied to Adam (termed AdamPower), GradPower consistently achieves lower terminal loss across diverse architectures (LLaMA, Qwen2MoE), parameter scales (66M to 2B), datasets (C4, OpenWebText), and learning-rate schedules (cosine, warmup-stable-decay). The most pronounced gains are observed when training modern mixture-of-experts models with warmup-stable-decay schedules. GradPower also integrates seamlessly with other state-of-the-art optimizers, such as Muon, yielding further improvements. Finally, we provide theoretical analyses that reveal the underlying mechanism of GradPower and highlight the influence of gradient noise.
Fonte: arXiv stat.ML
MLOps/Systems • Score 85
Model Merging via Data-Free Covariance Estimation
arXiv:2604.01329v1 Announce Type: new
Abstract: Model merging provides a way of cheaply combining individual models to produce a model that inherits each individual's capabilities. While some merging methods can approach the performance of multitask training, they are often heuristically motivated and lack theoretical justification. A principled alternative is to pose model merging as a layer-wise optimization problem that directly minimizes interference between tasks. However, this formulation requires estimating per-layer covariance matrices from data, which may not be available when performing merging. In contrast, many of the heuristically-motivated methods do not require auxiliary data, making them practically advantageous. In this work, we revisit the interference minimization framework and show that, under certain conditions, covariance matrices can be estimated directly from difference matrices, eliminating the need for data while also reducing computational costs. We validate our approach across vision and language benchmarks on models ranging from 86M parameters to 7B parameters, outperforming previous data-free state-of-the-art merging methods
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
SECURE: Stable Early Collision Understanding via Robust Embeddings in Autonomous Driving
arXiv:2604.01337v1 Announce Type: new
Abstract: While deep learning has significantly advanced accident anticipation, the robustness of these safety-critical systems against real-world perturbations remains a major challenge. We reveal that state-of-the-art models like CRASH, despite their high performance, exhibit significant instability in predictions and latent representations when faced with minor input perturbations, posing serious reliability risks. To address this, we introduce SECURE - Stable Early Collision Understanding Robust Embeddings, a framework that formally defines and enforces model robustness. SECURE is founded on four key attributes: consistency and stability in both prediction space and latent feature space. We propose a principled training methodology that fine-tunes a baseline model using a multi-objective loss, which minimizes divergence from a reference model and penalizes sensitivity to adversarial perturbations. Experiments on DAD and CCD datasets demonstrate that our approach not only significantly enhances robustness against various perturbations but also improves performance on clean data, achieving new state-of-the-art results.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Generative Profiling for Soft Real-Time Systems and its Applications to Resource Allocation
arXiv:2604.01441v1 Announce Type: cross
Abstract: Modern real-time systems require accurate characterization of task timing behavior to ensure predictable performance, particularly on complex hardware architectures. Existing methods, such as worst-case execution time analysis, often fail to capture the fine-grained timing behaviors of a task under varying resource contexts (e.g., an allocation of cache, memory bandwidth, and CPU frequency), which is necessary to achieve efficient resource utilization. In this paper, we introduce a novel generative profiling approach that synthesizes context-dependent, fine-grained timing profiles for real-time tasks, including those for unmeasured resource allocations. Our approach leverages a nonparametric, conditional multi-marginal Schr\"odinger Bridge (MSB) formulation to generate accurate execution profiles for unseen resource contexts, with maximum likelihood guarantees. We demonstrate the efficiency and effectiveness of our approach through real-world benchmarks, and showcase its practical utility in a representative case study of adaptive multicore resource allocation for real-time systems.
Fonte: arXiv stat.ML
Theory/Optimization • Score 85
Koopman Subspace Pruning in Reproducing Kernel Hilbert Spaces via Principal Vectors
arXiv:2604.01459v1 Announce Type: cross
Abstract: Data-driven approximations of the infinite-dimensional Koopman operator rely on finite-dimensional projections, where the predictive accuracy of the resulting models hinges heavily on the invariance of the chosen subspace. Subspace pruning systematically discards geometrically misaligned directions to enhance this invariance proximity, which formally corresponds to the largest principal angle between the subspace and its image under the operator. Yet, existing techniques are largely restricted to Euclidean settings. To bridge this gap, this paper presents an approach for computing principal angles and vectors to enable Koopman subspace pruning within a Reproducing Kernel Hilbert Space (RKHS) geometry. We first outline an exact computational routine, which is subsequently scaled for large datasets using randomized Nystrom approximations. Based on these foundations, we introduce the Kernel-SPV and Approximate Kernel-SPV algorithms for targeted subspace refinement via principal vectors. Simulation results validate our approach.
Fonte: arXiv stat.ML
RL • Score 85
Physics Informed Reinforcement Learning with Gibbs Priors for Topology Control in Power Grids
arXiv:2604.01830v1 Announce Type: new
Abstract: Topology control for power grid operation is a challenging sequential decision making problem because the action space grows combinatorially with the size of the grid and action evaluation through simulation is computationally expensive. We propose a physics-informed Reinforcement Learning framework that combines semi-Markov control with a Gibbs prior, that encodes the system's physics, over the action space. The decision is only taken when the grid enters a hazardous regime, while a graph neural network surrogate predicts the post action overload risk of feasible topology actions. These predictions are used to construct a physics-informed Gibbs prior that both selects a small state-dependent candidate set and reweights policy logits before action selection. In this way, our method reduces exploration difficulty and online simulation cost while preserving the flexibility of a learned policy. We evaluate the approach in three realistic benchmark environments of increasing difficulty. Across all settings, the proposed method achieves a strong balance between control quality and computational efficiency: it matches oracle-level performance while being approximately $6\times$ faster on the first benchmark, reaches $94.6\%$ of oracle reward with roughly $200\times$ lower decision time on the second one, and on the most challenging benchmark improves over a PPO baseline by up to $255\%$ in reward and $284\%$ in survived steps while remaining about $2.5\times$ faster than a strong specialized engineering baseline. These results show that our method provides an effective mechanism for topology control in power grids.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
PI-JEPA: Label-Free Surrogate Pretraining for Coupled Multiphysics Simulation via Operator-Split Latent Prediction
arXiv:2604.01349v1 Announce Type: new
Abstract: Reservoir simulation workflows face a fundamental data asymmetry: input parameter fields (geostatistical permeability realizations, porosity distributions) are free to generate in arbitrary quantities, yet existing neural operator surrogates require large corpora of expensive labeled simulation trajectories and cannot exploit this unlabeled structure. We introduce \textbf{PI-JEPA} (Physics-Informed Joint Embedding Predictive Architecture), a surrogate pretraining framework that trains \emph{without any completed PDE solves}, using masked latent prediction on unlabeled parameter fields under per-sub-operator PDE residual regularization. The predictor bank is structurally aligned with the Lie--Trotter operator-splitting decomposition of the governing equations, dedicating a separate physics-constrained latent module to each sub-process (pressure, saturation transport, reaction), enabling fine-tuning with as few as 100 labeled simulation runs. On single-phase Darcy flow, PI-JEPA achieves $1.9\times$ lower error than FNO and $2.4\times$ lower error than DeepONet at $N_\ell{=}100$, with 24\% improvement over supervised-only training at $N_\ell{=}500$, demonstrating that label-free surrogate pretraining substantially reduces the simulation budget required for multiphysics surrogate deployment.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Massively Parallel Exact Inference for Hawkes Processes
arXiv:2604.01342v1 Announce Type: new
Abstract: Multivariate Hawkes processes are a widely used class of self-exciting point processes, but maximum likelihood estimation naively scales as $O(N^2)$ in the number of events. The canonical linear exponential Hawkes process admits a faster $O(N)$ recurrence, but prior work evaluates this recurrence sequentially, without exploiting parallelization on modern GPUs. We show that the Hawkes process intensity can be expressed as a product of sparse transition matrices admitting a linear-time associative multiply, enabling computation via a parallel prefix scan. This yields a simple yet massively parallelizable algorithm for maximum likelihood estimation of linear exponential Hawkes processes. Our method reduces the computational complexity to approximately $O(N/P)$ with $P$ parallel processors, and naturally yields a batching scheme to maintain constant memory usage, avoiding GPU memory constraints. Importantly, it computes the exact likelihood without any additional assumptions or approximations, preserving the simplicity and interpretability of the model. We demonstrate orders-of-magnitude speedups on simulated and real datasets, scaling to thousands of nodes and tens of millions of events, substantially beyond scales reported in prior work. We provide an open-source PyTorch library implementing our optimizations.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
An Online Machine Learning Multi-resolution Optimization Framework for Energy System Design Limit of Performance Analysis
arXiv:2604.01308v1 Announce Type: new
Abstract: Designing reliable integrated energy systems for industrial processes requires optimization and verification models across multiple fidelities, from architecture-level sizing to high-fidelity dynamic operation. However, model mismatch across fidelities obscures the sources of performance loss and complicates the quantification of architecture-to-operation performance gaps. We propose an online, machine-learning-accelerated multi-resolution optimization framework that estimates an architecture-specific upper bound on achievable performance while minimizing expensive high-fidelity model evaluations. We demonstrate the approach on a pilot energy system supplying a 1 MW industrial heat load. First, we solve a multi-objective architecture optimization to select the system configuration and component capacities. We then develop an machine learning (ML)-accelerated multi-resolution, receding-horizon optimal control strategy that approaches the achievable-performance bound for the specified architecture, given the additional controls and dynamics not captured by the architectural optimization model. The ML-guided controller adaptively schedules the optimization resolution based on predictive uncertainty and warm-starts high-fidelity solves using elite low-fidelity solutions. Our results on the pilot case study show that the proposed multi-resolution strategy reduces the architecture-to-operation performance gap by up to 42% relative to a rule-based controller, while reducing required high-fidelity model evaluations by 34% relative to the same multi-fidelity approach without ML guidance, enabling faster and more reliable design verification. Together, these gains make high-fidelity verification tractable, providing a practical upper bound on achievable operational performance.
Fonte: arXiv cs.LG
RecSys • Score 85
Preference Guided Iterated Pareto Referent Optimisation for Accessible Route Planning
arXiv:2604.00795v1 Announce Type: new
Abstract: We propose the Preference Guided Iterated Pareto Referent Optimisation (PG-IPRO) for urban route planning for people with different accessibility requirements and preferences. With this algorithm the user can interact with the system by giving feedback on a route, i.e., the user can say which objective should be further minimized, or conversely can be relaxed. This leads to intuitive user interaction, that is especially effective during early iterations compared to information-gain-based interaction. Furthermore, due to PG-IPRO's iterative nature, the full set of alternative, possibly optimal policies (the Pareto front), is never computed, leading to higher computational efficiency and shorter waiting times for users.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Task-Centric Personalized Federated Fine-Tuning of Language Models
arXiv:2604.00050v1 Announce Type: new
Abstract: Federated Learning (FL) has emerged as a promising technique for training language models on distributed and private datasets of diverse tasks. However, aggregating models trained on heterogeneous tasks often degrades the overall performance of individual clients. To address this issue, Personalized FL (pFL) aims to create models tailored for each client's data distribution. Although these approaches improve local performance, they usually lack robustness in two aspects: (i) generalization: when clients must make predictions on unseen tasks, or face changes in their data distributions, and (ii) intra-client tasks interference: when a single client's data contains multiple distributions that may interfere with each other during local training. To tackle these two challenges, we propose FedRouter, a clustering-based pFL that builds specialized models for each task rather than for each client. FedRouter uses adapters to personalize models by employing two clustering mechanisms to associate adapters with specific tasks. A local clustering that associate adapters with task data samples and a global one that associates similar adapters from different clients to construct task-centric personalized models. Additionally, we propose an evaluation router mechanism that routes test samples to the best adapter based on the created clusters. Experiments comparing our method with existing approaches across a multitask dataset, FedRouter demonstrate strong resilience in these challenging scenarios performing up to 6.1% relatively better under tasks interference and up to 136% relative improvement under generalization evaluation.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Speeding Up Mixed-Integer Programming Solvers with Sparse Learning for Branching
arXiv:2604.00094v1 Announce Type: new
Abstract: Machine learning is increasingly used to improve decisions within branch-and-bound algorithms for mixed-integer programming. Many existing approaches rely on deep learning, which often requires very large training datasets and substantial computational resources for both training and deployment, typically with GPU parallelization. In this work, we take a different path by developing interpretable models that are simple but effective. We focus on approximating strong branching (SB) scores, a highly effective yet computationally expensive branching rule. Using sparse learning methods, we build models with fewer than 4% of the parameters of a state-of-the-art graph neural network (GNN) while achieving competitive accuracy. Relative to SCIP's built-in branching rules and the GNN-based model, our CPU-only models are faster than the default solver and the GPU-accelerated GNN. The models are simple to train and deploy, and they remain effective with small training sets, which makes them practical in low-resource settings. Extensive experiments across diverse problem classes demonstrate the efficiency of this approach.
Fonte: arXiv cs.LG
NLP/LLMs • Score 95
MAC-Attention: a Match-Amend-Complete Scheme for Fast and Accurate Attention Computation
arXiv:2604.00235v1 Announce Type: new
Abstract: Long-context decoding in LLMs is IO-bound: each token re-reads an ever-growing KV cache. Prior accelerations cut bytes via compression, which lowers fidelity, or selection/eviction, which restricts what remains accessible, and both can degrade delayed recall and long-form generation. We introduce MAC-Attention, a fidelity- and access-preserving alternative that accelerates decoding by reusing prior attention computations for semantically similar recent queries. It starts with a match stage that performs pre-RoPE L2 matching over a short local window; an amend stage rectifies the reused attention by recomputing a small band near the match boundary; and a complete stage fuses the rectified results with fresh attention computed on the KV tail through a numerically stable merge. On a match hit, the compute and bandwidth complexity is constant regardless of context length. The method is model-agnostic and composes with IO-aware kernels, paged-KV managers, and MQA/GQA. Across LongBench v2 (120K), RULER (120K), and LongGenBench (16K continuous generation), compared to the latest FlashInfer library, MAC-Attention reduces KV accesses by up to 99%, cuts token generation latency by over 60% at 128K, and achieves over 14.3x attention-phase speedups, up to 2.6x end-to-end, while maintaining full-attention quality. By reusing computation, MAC-Attention delivers long-context inference that is both fast and faithful. Code is available here: https://github.com/YJHMITWEB/MAC-Attention.git
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
BloClaw: An Omniscient, Multi-Modal Agentic Workspace for Next-Generation Scientific Discovery
arXiv:2604.00550v1 Announce Type: new
Abstract: The integration of Large Language Models (LLMs) into life sciences has catalyzed the development of "AI Scientists." However, translating these theoretical capabilities into deployment-ready research environments exposes profound infrastructural vulnerabilities. Current frameworks are bottlenecked by fragile JSON-based tool-calling protocols, easily disrupted execution sandboxes that lose graphical outputs, and rigid conversational interfaces inherently ill-suited for high-dimensional scientific data.We introduce BloClaw, a unified, multi-modal operating system designed for Artificial Intelligence for Science (AI4S). BloClaw reconstructs the Agent-Computer Interaction (ACI) paradigm through three architectural innovations: (1) An XML-Regex Dual-Track Routing Protocol that statistically eliminates serialization failures (0.2% error rate vs. 17.6% in JSON); (2) A Runtime State Interception Sandbox that utilizes Python monkey-patching to autonomously capture and compile dynamic data visualizations (Plotly/Matplotlib), circumventing browser CORS policies; and (3) A State-Driven Dynamic Viewport UI that morphs seamlessly between a minimalist command deck and an interactive spatial rendering engine. We comprehensively benchmark BloClaw across cheminformatics (RDKit), de novo 3D protein folding via ESMFold, molecular docking, and autonomous Retrieval-Augmented Generation (RAG), establishing a highly robust, self-evolving paradigm for computational research assistants. The open-source repository is available at https://github.com/qinheming/BloClaw.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Decision-Centric Design for LLM Systems
arXiv:2604.00414v1 Announce Type: new
Abstract: LLM systems must make control decisions in addition to generating outputs: whether to answer, clarify, retrieve, call tools, repair, or escalate. In many current architectures, these decisions remain implicit within generation, entangling assessment and action in a single model call and making failures hard to inspect, constrain, or repair. We propose a decision-centric framework that separates decision-relevant signals from the policy that maps them to actions, turning control into an explicit and inspectable layer of the system. This separation supports attribution of failures to signal estimation, decision policy, or execution, and enables modular improvement of each component. It unifies familiar single-step settings such as routing and adaptive inference, and extends naturally to sequential settings in which actions alter the information available before acting. Across three controlled experiments, the framework reduces futile actions, improves task success, and reveals interpretable failure modes. More broadly, it offers a general architectural principle for building more reliable, controllable, and diagnosable LLM systems.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Open, Reliable, and Collective: A Community-Driven Framework for Tool-Using AI Agents
arXiv:2604.00137v1 Announce Type: new
Abstract: Tool-integrated LLMs can retrieve, compute, and take real-world actions via external tools, but reliability remains a key bottleneck. We argue that failures stem from both tool-use accuracy (how well an agent invokes a tool) and intrinsic tool accuracy (the tool's own correctness), while most prior work emphasizes the former. We introduce OpenTools, a community-driven toolbox that standardizes tool schemas, provides lightweight plug-and-play wrappers, and evaluates tools with automated test suites and continuous monitoring. We also release a public web demo where users can run predefined agents and tools and contribute test cases, enabling reliability reports to evolve as tools change. OpenTools includes the core framework, an initial tool set, evaluation pipelines, and a contribution protocol. Experiments and evaluations show improved end-to-end reproducibility and task performance; community-contributed, higher-quality task-specific tools deliver 6%-22% relative gains over an existing toolbox across multiple agent architectures on downstream tasks and benchmarks, highlighting the importance of intrinsic tool accuracy.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Disentanglement of Sources in a Multi-Stream Variational Autoencoder
arXiv:2510.15669v2 Announce Type: replace
Abstract: Variational autoencoders (VAEs) are among leading approaches to address the problem of learning disentangled representations. Typically a single VAE is used and disentangled representations are sought within its single continuous latent space. In this paper, we propose and provide a proof of concept for a novel Multi-Stream Variational Autoencoder (MS-VAE) that achieves disentanglement of sources by combining discrete and continuous latents. The discrete latents are used in an explicit source combination model, that superimposes a set of sources as part of the MS-VAE decoder. We formally define the MS-VAE approach, derive its inference and learning equations, and numerically investigate its principled functionality. The MS-VAE model is very flexible and can be trained using little supervision (we use fully unsupervised learning after pretraining with some labels). In our numerical experiments, we explored the ability of the MS-VAE approach in separating both superimposed hand-written digits as well as sound sources. For the former task we used superimposed MNIST digits (an increasingly common benchmark). For sound separation, our experiments focused on the task of speaker diarization in a recording conversation between two speakers. In all cases, we observe a clear separation of sources and competitive performance after training. For digit superpositions, performance is particularly competitive in complex mixtures (e.g., three and four digits). For the speaker diarization task, we observe an especially low rate of missed speakers and a more precise speaker attribution. Numerical experiments confirm the flexibility of the approach across varying amounts of supervision, and we observed high performance, e.g., when using just 10% of the labels for pretraining.
Fonte: arXiv stat.ML
Theory/Optimization • Score 85
Natural Hypergradient Descent: Algorithm Design, Convergence Analysis, and Parallel Implementation
arXiv:2602.10905v2 Announce Type: replace-cross
Abstract: In this work, we propose Natural Hypergradient Descent (NHGD), a new method for solving bilevel optimization problems. To address the computational bottleneck in hypergradient estimation--namely, the need to compute or approximate Hessian inverse--we exploit the statistical structure of the inner optimization problem and use the empirical Fisher information matrix as an asymptotically consistent surrogate for the Hessian. This design enables a parallel optimize-and-approximate framework in which the Hessian-inverse approximation is updated synchronously with the stochastic inner optimization, reusing gradient information at negligible additional cost. Our main theoretical contribution establishes high-probability error bounds and sample complexity guarantees for NHGD that match those of state-of-the-art optimize-then-approximate methods, while significantly reducing computational time overhead. Empirical evaluations on representative bilevel learning tasks further demonstrate the practical advantages of NHGD, highlighting its scalability and effectiveness in large-scale machine learning settings.
Fonte: arXiv stat.ML
Multimodal • Score 90
OmniMem: Autoresearch-Guided Discovery of Lifelong Multimodal Agent Memory
arXiv:2604.01007v1 Announce Type: new
Abstract: AI agents increasingly operate over extended time horizons, yet their ability to retain, organize, and recall multimodal experiences remains a critical bottleneck. Building effective lifelong memory requires navigating a vast design space spanning architecture, retrieval strategies, prompt engineering, and data pipelines; this space is too large and interconnected for manual exploration or traditional AutoML to explore effectively. We deploy an autonomous research pipeline to discover OmniMem, a unified multimodal memory framework for lifelong AI agents. Starting from a na\"ive baseline (F1=0.117 on LoCoMo), the pipeline autonomously executes ${\sim}50$ experiments across two benchmarks, diagnosing failure modes, proposing architectural modifications, and repairing data pipeline bugs, all without human intervention in the inner loop. The resulting system achieves state-of-the-art on both benchmarks, improving F1 by +411% on LoCoMo (0.117$\to$0.598) and +214% on Mem-Gallery (0.254$\to$0.797) relative to the initial configurations. Critically, the most impactful discoveries are not hyperparameter adjustments: bug fixes (+175%), architectural changes (+44%), and prompt engineering (+188\% on specific categories) each individually exceed the cumulative contribution of all hyperparameter tuning, demonstrating capabilities fundamentally beyond the reach of traditional AutoML. We provide a taxonomy of six discovery types and identify four properties that make multimodal memory particularly suited for autoresearch, offering guidance for applying autonomous research pipelines to other AI system domains. Code is available at this https://github.com/aiming-lab/OmniMem.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
Learning Hyperparameters via a Data-Emphasized Variational Objective
arXiv:2502.01861v3 Announce Type: replace-cross
Abstract: When training large models on limited data, avoiding overfitting is paramount. Common grid search or smarter search methods rely on expensive separate runs for each candidate hyperparameter, while carving out a validation set that reduces available training data. In this paper, we study gradient-based learning of hyperparameters via the evidence lower bound (ELBO) objective from Bayesian variational methods. This avoids the need for any validation set. We focus on scenarios where the model is over-parameterized for flexibility and the approximate posterior is chosen to be Gaussian with isotropic covariance for tractability, even though it cannot match the true posterior. In such scenarios, we find the ELBO prioritizes posteriors that match the prior, leading to severe underfitting. Instead, we recommend a data-emphasized ELBO that upweights the likelihood but not the prior. In Bayesian transfer learning of image and text classifiers, our method reduces the 88+ hour grid search of past work to under 3 hours while delivering comparable accuracy. We further demonstrate how our approach enables efficient yet accurate approximations of Gaussian processes with learnable lengthscale kernels.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
More Human, More Efficient: Aligning Annotations with Quantized SLMs
arXiv:2604.00586v1 Announce Type: new
Abstract: As Large Language Model (LLM) capabilities advance, the demand for high-quality annotation of exponentially increasing text corpora has outpaced human capacity, leading to the widespread adoption of LLMs in automatic evaluation and annotation. However, proprietary LLMs often exhibit systematic biases that diverge from human expert consensus, lacks reproducibility, and raises data privacy concerns. Our work examines the viability of finetuning a quantized Small Language Model of 1.7B parameter size on limited human-annotated data to serve as a highly aligned, deterministic evaluator and annotator. By implementing a custom, multi-dimensional rubric framework and simple augmentation and regularization techniques, the proposed approach achieves higher inter-annotator agreement (0.23 points increase in Krippendorff's $\alpha$) than the best performing state-of-the-art proprietary LLM. We also demonstrate the generalizability of the proposed training pipeline on a separate emotion classification task. The results show that task-specific alignment and efficient 4-bit quantized fine-tuning provide superior open-source alternative to using proprietary models for evaluation and annotation. Our finetuning approach is publicly available at https://github.com/jylee-k/slm-judge.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
Optimsyn: Influence-Guided Rubrics Optimization for Synthetic Data Generation
arXiv:2604.00536v1 Announce Type: new
Abstract: Large language models (LLMs) achieve strong downstream performance largely due to abundant supervised fine-tuning (SFT) data. However, high-quality SFT data in knowledge-intensive domains such as humanities, social sciences, medicine, law, and finance is scarce because expert curation is expensive, privacy constraints are strict, and label consistency is hard to ensure. Recent work uses synthetic data, typically by prompting a generator over domain documents and filtering outputs with handcrafted rubrics. Yet rubric design is expert-dependent, transfers poorly across domains, and is often optimized through a brittle heuristic loop of writing rubrics, synthesizing data, training, inspecting results, and manually guessing revisions. This process lacks reliable quantitative feedback about how a rubric affects downstream performance. We propose evaluating synthetic data by its training utility on the target model and using this signal to guide data generation. Inspired by influence estimation, we adopt an optimizer-aware estimator that uses gradient information to quantify each synthetic sample's contribution to a target model's objective on specific tasks. Our analysis shows that even when synthetic and real samples are close in embedding space, their influence on learning can differ substantially. Based on this insight, we propose an optimization-based framework that adapts rubrics using target-model feedback. We provide lightweight guiding text and use a rubric-specialized model to generate task-conditioned rubrics. Influence score is used as the reward to optimize the rubric generator with reinforcement learning. Experiments across domains, target models, and data generators show consistent improvements and strong generalization without task-specific tuning.
Fonte: arXiv cs.CL
Theory/Optimization • Score 85
Isomorphic Functionalities between Ant Colony and Ensemble Learning: Part II-On the Strength of Weak Learnability and the Boosting Paradigm
arXiv:2604.00038v1 Announce Type: new
Abstract: In Part I of this series, we established a rigorous mathematical isomorphism between ant colony decision-making and random forest learning, demonstrating that variance reduction through decorrelation is a universal principle shared by biological and computational ensembles. Here we turn to the complementary mechanism: bias reduction through adaptive weighting. Just as boosting algorithms sequentially focus on difficult instances, ant colonies dynamically amplify successful foraging paths through pheromone-mediated recruitment. We prove that these processes are mathematically isomorphic, establishing that the fundamental theorem of weak learnability has a direct analog in colony decision-making. We develop a formal mapping between AdaBoost's adaptive reweighting and ant recruitment dynamics, show that the margin theory of boosting corresponds to the stability of quorum decisions, and demonstrate through comprehensive simulation that ant colonies implementing adaptive recruitment achieve the same bias-reduction benefits as boosting algorithms. This completes a unified theory of ensemble intelligence, revealing that both variance reduction (Part I) and bias reduction (Part II) are manifestations of the same underlying mathematical principles governing collective intelligence in biological and computational systems.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Eyla: Toward an Identity-Anchored LLM Architecture with Integrated Biological Priors -- Vision, Implementation Attempt, and Lessons from AI-Assisted Development
arXiv:2604.00009v1 Announce Type: cross
Abstract: We present the design rationale, implementation attempt, and failure analysis of Eyla, a proposed identity-anchored LLM architecture that integrates biologically-inspired subsystems -- including HiPPO-initialized state-space models, zero-initialized adapters, episodic memory retrieval, and calibrated uncertainty training -- into a unified agent operating system running on consumer hardware. Unlike existing approaches that optimize models for generic helpfulness, Eyla targets identity consistency: the ability to maintain a coherent self-model under adversarial pressure, admit uncertainty, and resist manipulation. We propose the Identity Consistency Score (ICS), a novel benchmark for evaluating this property across LLMs. We then present an honest account of attempting to implement this architecture using AI coding assistants (Claude Code, Cursor) as a non-programmer, documenting a $1,000+ failure that produced a 1.27B parameter model with 86 brain subsystems contributing less than 2% to output. Our analysis identifies five systematic failure modes of AI-assisted development for novel architectures and offers concrete recommendations. To our knowledge, this is the first paper to combine an architectural vision with a documented first-person failure analysis of AI-assisted LLM development, providing lessons for both the AI systems and AI-assisted software engineering communities.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
MF-QAT: Multi-Format Quantization-Aware Training for Elastic Inference
arXiv:2604.00529v1 Announce Type: new
Abstract: Quantization-aware training (QAT) is typically performed for a single target numeric format, while practical deployments often need to choose numerical precision at inference time based on hardware support or runtime constraints. We study multi-format QAT, where a single model is trained to be robust across multiple quantization formats. We find that multi-format QAT can match single-format QAT at each target precision, yielding one model that performs well overall across different formats, even formats that were not seen during training. To enable practical deployment, we propose the Slice-and-Scale conversion procedure for both MXINT and MXFP that converts a high-precision representation into lower-precision formats without re-training. Building on this, we introduce a pipeline that (i) trains a model with multi-format QAT, (ii) stores a single anchor format checkpoint (MXINT8/MXFP8), and (iii) allows on-the-fly conversion to lower MXINT or MXFP formats at runtime with negligible-or no-additional accuracy degradation. Together, these components provide a practical path to elastic precision scaling and allow selecting the runtime format at inference time across diverse deployment targets.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Convergence of Byzantine-Resilient Gradient Tracking via Probabilistic Edge Dropout
arXiv:2604.00449v1 Announce Type: new
Abstract: We study distributed optimization over networks with Byzantine agents that may send arbitrary adversarial messages. We propose \emph{Gradient Tracking with Probabilistic Edge Dropout} (GT-PD), a stochastic gradient tracking method that preserves the convergence properties of gradient tracking under adversarial communication. GT-PD combines two complementary defense layers: a universal self-centered projection that clips each incoming message to a ball of radius $\tau$ around the receiving agent, and a fully decentralized probabilistic dropout rule driven by a dual-metric trust score in the decision and tracking channels. This design bounds adversarial perturbations while preserving the doubly stochastic mixing structure, a property often lost under robust aggregation in decentralized settings. Under complete Byzantine isolation ($p_b=0$), GT-PD converges linearly to a neighborhood determined solely by stochastic gradient variance. For partial isolation ($p_b>0$), we introduce \emph{Gradient Tracking with Probabilistic Edge Dropout and Leaky Integration} (GT-PD-L), which uses a leaky integrator to control the accumulation of tracking errors caused by persistent perturbations and achieves linear convergence to a bounded neighborhood determined by the stochastic variance and the clipping-to-leak ratio. We further show that under two-tier dropout with $p_h=1$, isolating Byzantine agents introduces no additional variance into the honest consensus dynamics. Experiments on MNIST under Sign Flip, ALIE, and Inner Product Manipulation attacks show that GT-PD-L outperforms coordinate-wise trimmed mean by up to 4.3 percentage points under stealth attacks.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Self-Routing: Parameter-Free Expert Routing from Hidden States
arXiv:2604.00421v1 Announce Type: new
Abstract: Mixture-of-Experts (MoE) layers increase model capacity by activating only a small subset of experts per token, and typically rely on a learned router to map hidden states to expert assignments. In this work, we ask whether a dedicated learned router is strictly necessary in the MoE settings we study. We propose Self-Routing, a parameter-free routing mechanism that uses a designated subspace of the token hidden state directly as expert logits, eliminating the router projection entirely while leaving the rest of the MoE layer unchanged. We evaluate Self-Routing on GPT-2-scale language modeling and ImageNet-1K classification by comparing it against a standard learned router, random-routing baselines, and dense non-MoE baselines. Our results show that Self-Routing remains competitive with the learned-router baseline while removing all dedicated routing parameters, and yields more balanced expert utilization, with about 17 % higher average normalized routing entropy and no explicit load-balancing loss. On ImageNet-1K with DeiT-S/16, Self-Routing also slightly improves over the corresponding learned-router MoE. These findings suggest that effective MoE routing can emerge from the hidden representation itself without requiring a separate learned router module.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Oblivion: Self-Adaptive Agentic Memory Control through Decay-Driven Activation
arXiv:2604.00131v1 Announce Type: new
Abstract: Human memory adapts through selective forgetting: experiences become less accessible over time but can be reactivated by reinforcement or contextual cues. In contrast, memory-augmented LLM agents rely on "always-on" retrieval and "flat" memory storage, causing high interference and latency as histories grow. We introduce Oblivion, a memory control framework that casts forgetting as decay-driven reductions in accessibility, not explicit deletion. Oblivion decouples memory control into read and write paths. The read path decides when to consult memory, based on agent uncertainty and memory buffer sufficiency, avoiding redundant always-on access. The write path decides what to strengthen, by reinforcing memories contributing to forming the response. Together, this enables hierarchical memory organization that maintains persistent high-level strategies while dynamically loading details as needed. We evaluate on both static and dynamic long-horizon interaction benchmarks. Results show that Oblivion dynamically adapts memory access and reinforcement, balancing learning and forgetting under shifting contexts, highlighting that memory control is essential for effective LLM-agentic reasoning. The source code is available at https://github.com/nec-research/oblivion.
Fonte: arXiv cs.CL
Theory/Optimization • Score 85
Inverse-Free Sparse Variational Gaussian Processes
arXiv:2604.00697v1 Announce Type: new
Abstract: Gaussian processes (GPs) offer appealing properties but are costly to train at scale. Sparse variational GP (SVGP) approximations reduce cost yet still rely on Cholesky decompositions of kernel matrices, ill-suited to low-precision, massively parallel hardware. While one can construct valid variational bounds that rely only on matrix multiplications (matmuls) via an auxiliary matrix parameter, optimising them with off-the-shelf first-order methods is challenging. We make the inverse-free approach practical by proposing a better-conditioned bound and deriving a matmul-only natural-gradient update for the auxiliary parameter, markedly improving stability and convergence. We further provide simple heuristics, such as step-size schedules and stopping criteria, that make the overall optimisation routine fit seamlessly into existing workflows. Across regression and classification benchmarks, we demonstrate that our method 1) serves as a drop-in replacement in SVGP-based models (e.g., deep GPs), 2) recovers similar performance to traditional methods, and 3) can be faster than baselines when well tuned.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Adaptive Parallel Monte Carlo Tree Search for Efficient Test-time Compute Scaling
arXiv:2604.00510v1 Announce Type: new
Abstract: Monte Carlo Tree Search (MCTS) is an effective test-time compute scaling (TTCS) method for improving the reasoning performance of large language models, but its highly variable execution time leads to severe long-tail latency in practice. Existing optimizations such as positive early exit, reduce latency in favorable cases but are less effective when search continues without meaningful progress. We introduce {\it negative early exit}, which prunes unproductive MCTS trajectories, and an {\it adaptive boosting mechanism} that reallocates reclaimed computation to reduce resource contention among concurrent searches. Integrated into vLLM, these techniques substantially reduce p99 end-to-end latency while improving throughput and maintaining reasoning accuracy.
Fonte: arXiv cs.AI
Multimodal • Score 85
MOON3.0: Reasoning-aware Multimodal Representation Learning for E-commerce Product Understanding
arXiv:2604.00513v1 Announce Type: new
Abstract: With the rapid growth of e-commerce, exploring general representations rather than task-specific ones has attracted increasing attention. Although recent multimodal large language models (MLLMs) have driven significant progress in product understanding, they are typically employed as feature extractors that implicitly encode product information into global embeddings, thereby limiting their ability to capture fine-grained attributes. Therefore, we argue that leveraging the reasoning capabilities of MLLMs to explicitly model fine-grained product attributes holds significant potential. Nevertheless, achieving this goal remains non-trivial due to several key challenges: (i) long-context reasoning tends to dilute the model's attention to salient information in the raw input; (ii) supervised fine-tuning (SFT) primarily encourages rigid imitation, limiting the exploration of effective reasoning strategies; and (iii) fine-grained details are progressively attenuated during forward propagation. To address these issues, we propose MOON3.0, the first reasoning-aware MLLM-based model for product representation learning. Our method (1) employs a multi-head modality fusion module to adaptively integrate raw signals; (2) incorporates a joint contrastive and reinforcement learning framework to autonomously explore more effective reasoning strategies; and (3) introduces a fine-grained residual enhancement module to progressively preserve local details throughout the network. Additionally, we release a large-scale multimodal e-commerce benchmark MBE3.0. Experimentally, our model demonstrates state-of-the-art zero-shot performance across various downstream tasks on both our benchmark and public datasets.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Collaborative AI Agents and Critics for Fault Detection and Cause Analysis in Network Telemetry
arXiv:2604.00319v1 Announce Type: new
Abstract: We develop algorithms for collaborative control of AI agents and critics in a multi-actor, multi-critic federated multi-agent system. Each AI agent and critic has access to classical machine learning or generative AI foundation models. The AI agents and critics collaborate with a central server to complete multimodal tasks such as fault detection, severity, and cause analysis in a network telemetry system, text-to-image generation, video generation, healthcare diagnostics from medical images and patient records, etcetera. The AI agents complete their tasks and send them to AI critics for evaluation. The critics then send feedback to agents to improve their responses. Collaboratively, they minimize the overall cost to the system with no inter-agent or inter-critic communication. AI agents and critics keep their cost functions or derivatives of cost functions private. Using multi-time scale stochastic approximation techniques, we provide convergence guarantees on the time-average active states of AI agents and critics. The communication overhead is a little on the system, of the order of $\mathcal{O}(m)$, for $m$ modalities and is independent of the number of AI agents and critics. Finally, we present an example of fault detection, severity, and cause analysis in network telemetry and thorough evaluation to check the algorithm's efficacy.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
SYNTHONY: A Stress-Aware, Intent-Conditioned Agent for Deep Tabular Generative Models Selection
arXiv:2604.00293v1 Announce Type: new
Abstract: Deep generative models for tabular data (GANs, diffusion models, and LLM-based generators) exhibit highly non-uniform behavior across datasets; the best-performing synthesizer family depends strongly on distributional stressors such as long-tailed marginals, high-cardinality categorical, Zipfian imbalance, and small-sample regimes. This brittleness makes practical deployment challenging, especially when users must balance competing objectives of fidelity, privacy, and utility. We study {intent-conditioned tabular synthesis selection}: given a dataset and a user intent expressed as a preference over evaluation metrics, the goal is to select a synthesizer that minimizes regret relative to an intent-specific oracle. We propose {stress profiling}, a synthesis-specific meta-feature representation that quantifies dataset difficulty along four interpretable stress dimensions, and integrate it into {SYNTHONY}, a selection framework that matches stress profiles against a calibrated capability registry of synthesizer families. Across a benchmark of 7 datasets, 10 synthesizers, and 3 intents, we demonstrate that stress-based meta-features are highly predictive of synthesizer performance: a $k$NN selector using these features achieves strong Top-1 selection accuracy, substantially outperforming zero-shot LLM selectors and random baselines. We analyze the gap between meta-feature-based and capability-based selection, identifying the hand-crafted capability registry as the primary bottleneck and motivating learned capability representations as a direction for future work.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
ParetoBandit: Budget-Paced Adaptive Routing for Non-Stationary LLM Serving
arXiv:2604.00136v1 Announce Type: new
Abstract: Production LLM serving often relies on multi-model portfolios spanning a ~530x cost range, where routing decisions trade off quality against cost. This trade-off is non-stationary: providers revise pricing, model quality can regress silently, and new models must be integrated without downtime. We present ParetoBandit, an open-source adaptive router built on cost-aware contextual bandits that is the first to simultaneously enforce dollar-denominated budgets, adapt online to such shifts, and onboard new models at runtime.
ParetoBandit closes these gaps through three mechanisms. An online primal-dual budget pacer enforces a per-request cost ceiling over an open-ended stream, replacing offline penalty tuning with closed-loop control. Geometric forgetting on sufficient statistics enables rapid adaptation to price and quality shifts while bootstrapping from offline priors. A hot-swap registry lets operators add or remove models at runtime, with a brief forced-exploration phase for each newcomer, after which UCB selection discovers its quality-cost niche from live traffic alone.
We evaluate ParetoBandit across four deployment scenarios on 1,824 prompts routed through a three-model portfolio. Across seven budget ceilings, mean per-request cost never exceeds the target by more than 0.4%. When conditions shift, the system adapts: an order-of-magnitude price cut on the costliest model yields up to +0.071 quality lift, and a silent quality regression is detected and rerouted within budget. A cold-started model reaches meaningful adoption within ~142 steps without breaching the cost ceiling. The router discriminates rather than blindly adopting: expensive models are budget-gated and low-quality models rejected after bounded exploration. End-to-end routing latency is 9.8ms on CPU -- less than 0.4% of typical inference time -- with the routing decision itself taking just 22.5us.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Deep Learning-Accelerated Surrogate Optimization for High-Dimensional Well Control in Stress-Sensitive Reservoirs
arXiv:2604.00352v1 Announce Type: new
Abstract: Production optimization in stress-sensitive unconventional reservoirs is governed by a nonlinear trade-off between pressure-driven flow and stress-induced degradation of fracture conductivity and matrix permeability. While higher drawdown improves short-term production, it accelerates permeability loss and reduces long-term recovery. Identifying optimal, time-varying control strategies requires repeated evaluations of fully coupled flow-geomechanics simulators, making conventional optimization computationally expensive.
We propose a deep learning-based surrogate optimization framework for high-dimensional well control. Unlike prior approaches that rely on predefined control parameterizations or generic sampling, our method treats well control as a continuous, high-dimensional problem and introduces a problem-informed sampling strategy that aligns training data with trajectories encountered during optimization. A neural network proxy is trained to approximate the mapping between bottomhole pressure trajectories and cumulative production using data from a coupled flow-geomechanics model.
The proxy is embedded within a constrained optimization workflow, enabling rapid evaluation of control strategies. Across multiple initializations, the surrogate achieves agreement with full-physics solutions within 2-5 percent, while reducing computational cost by up to three orders of magnitude. Discrepancies are mainly associated with trajectories near the boundary of the training distribution and local optimization effects.
This framework shows that combining surrogate modeling with problem-informed sampling enables scalable and reliable optimization for high-dimensional, simulator-based problems, with broader applicability to PDE-constrained systems.
Fonte: arXiv cs.LG
RL • Score 85
Execution-Verified Reinforcement Learning for Optimization Modeling
arXiv:2604.00442v1 Announce Type: new
Abstract: Automating optimization modeling with LLMs is a promising path toward scalable decision intelligence, but existing approaches either rely on agentic pipelines built on closed-source LLMs with high inference latency, or fine-tune smaller LLMs using costly process supervision that often overfits to a single solver API. Inspired by reinforcement learning with verifiable rewards, we propose Execution-Verified Optimization Modeling (EVOM), an execution-verified learning framework that treats a mathematical programming solver as a deterministic, interactive verifier. Given a natural-language problem and a target solver, EVOM generates solver-specific code, executes it in a sandboxed harness, and converts execution outcomes into scalar rewards, optimized with GRPO and DAPO in a closed-loop generate-execute-feedback-update process. This outcome-only formulation removes the need for process-level supervision, and enables cross-solver generalization by switching the verification environment rather than reconstructing solver-specific datasets. Experiments on NL4OPT, MAMO, IndustryOR, and OptiBench across Gurobi, OR-Tools, and COPT show that EVOM matches or outperforms process-supervised SFT, supports zero-shot solver transfer, and achieves effective low-cost solver adaptation by continuing training under the target solver backend.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
Neural Collapse Dynamics: Depth, Activation, Regularisation, and Feature Norm Threshold
arXiv:2604.00230v1 Announce Type: new
Abstract: Neural collapse (NC) -- the convergence of penultimate-layer features to a simplex equiangular tight frame -- is well understood at equilibrium, but the dynamics governing its onset remain poorly characterised. We identify a simple and predictive regularity: NC occurs when the mean feature norm reaches a model-dataset-specific critical value, fn*, that is largely invariant to training conditions. This value concentrates tightly within each (model, dataset) pair (CV 0.2). Completing the (architecture)x(dataset) grid reveals the paper's strongest result: ResNet-20 on MNIST gives fn* = 5.867 -- a +458% architecture effect versus only +68% on CIFAR-10. The grid is strongly non-additive; fn* cannot be decomposed into independent architecture and dataset contributions. Four structural regularities emerge: (1) depth has a non-monotonic effect on collapse speed; (2) activation jointly determines both collapse speed and fn*; (3) weight decay defines a three-regime phase diagram -- too little slows, an optimal range is fastest, and too much prevents collapse; (4) width monotonically accelerates collapse while shifting fn* by at most 13%. These results establish feature-norm dynamics as an actionable diagnostic for predicting NC timing, suggesting that norm-threshold behaviour is a general mechanism underlying delayed representational reorganisation in deep networks.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Scheduling LLM Inference with Uncertainty-Aware Output Length Predictions
arXiv:2604.00499v1 Announce Type: new
Abstract: To schedule LLM inference, the \textit{shortest job first} (SJF) principle is favorable by prioritizing requests with short output lengths to avoid head-of-line (HOL) blocking. Existing methods usually predict a single output length for each request to facilitate scheduling. We argue that such a \textit{point estimate} does not match the \textit{stochastic} decoding process of LLM inference, where output length is \textit{uncertain} by nature and determined by when the end-of-sequence (EOS) token is sampled. Hence, the output length of each request should be fitted with a distribution rather than a single value. With an in-depth analysis of empirical data and the stochastic decoding process, we observe that output length follows a heavy-tailed distribution and can be fitted with the log-t distribution. On this basis, we propose a simple metric called Tail Inflated Expectation (TIE) to replace the output length in SJF scheduling, which adjusts the expectation of a log-t distribution with its tail probabilities to account for the risk that a request generates long outputs. To evaluate our TIE scheduler, we compare it with three strong baselines, and the results show that TIE reduces the per-token latency by $2.31\times$ for online inference and improves throughput by $1.42\times$ for offline data generation.
Fonte: arXiv cs.LG
Applications • Score 85
Sit-to-Stand Transitions Detection and Duration Measurement Using Smart Lacelock Sensor
arXiv:2604.00175v1 Announce Type: new
Abstract: Postural stability during movement is fundamental to independent living, fall prevention, and overall health, particularly among older adults who experience age-related declines in balance, muscle strength, and mobility. Among daily functional activities, the Sit-to-Stand (SiSt) transition is a critical indicator of lower-limb strength, musculoskeletal health, and fall risk, making it an essential parameter for assessing functional capacity and monitoring physical decline in aging populations. This study presents a methodology SiSt transition detection and duration measurement using the Smart Lacelock sensor, a lightweight, shoe-mounted device that integrates a load cell, accelerometer, and gyroscope for motion analysis. The methodology was evaluated in 16 older adults (age: mean: 76.84, SD: 3.45 years) performing SiSt tasks within the Short Physical Performance Battery (SPPB) protocol. Features extracted from multimodal signals were used to train and evaluate four machine learning classifiers using a 4-fold participant-independent cross-validation to classify SiSt transitions and measure their duration. The bagged tree classifier achieved an accuracy of 0.98 and an F1 score of 0.8 in classifying SiSt transition. The mean absolute error in duration measurement of the correctly classified transitions was 0.047, and the SD was 0.07 seconds. These findings highlight the potential of the Smart Lacelock sensor for real-world fall-risk assessment and mobility monitoring in older adults.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Experience as a Compass: Multi-agent RAG with Evolving Orchestration and Agent Prompts
arXiv:2604.00901v1 Announce Type: new
Abstract: Multi-agent Retrieval-Augmented Generation (RAG), wherein each agent takes on a specific role, supports hard queries that require multiple steps and sources, or complex reasoning. Existing approaches, however, rely on static agent behaviors and fixed orchestration strategies, leading to brittle performance on diverse, multi-hop tasks. We identify two key limitations: the lack of continuously adaptive orchestration mechanisms and the absence of behavior-level learning for individual agents. To this end, we propose HERA, a hierarchical framework that jointly evolves multi-agent orchestration and role-specific agent prompts. At the global level, HERA optimizes query-specific agent topologies through reward-guided sampling and experience accumulation. At the local level, Role-Aware Prompt Evolution refines agent behaviors via credit assignment and dual-axes adaptation along operational and behavioral principles, enabling targeted, role-conditioned improvements. On six knowledge-intensive benchmarks, HERA achieves an average improvement of 38.69\% over recent baselines while maintaining robust generalization and token efficiency. Topological analyses reveal emergent self-organization, where sparse exploration yields compact, high-utility multi-agent networks, demonstrating both efficient coordination and robust reasoning.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Two-Stage Optimizer-Aware Online Data Selection for Large Language Models
arXiv:2604.00001v1 Announce Type: cross
Abstract: Gradient-based data selection offers a principled framework for estimating sample utility in large language model (LLM) fine-tuning, but existing methods are mostly designed for offline settings. They are therefore less suited to online fine-tuning, where data arrives sequentially, sample utility is step-dependent, and the effective update geometry is shaped by adaptive optimizers. We propose an optimizer-aware framework for gradient-based online data selection and reweighting in LLM fine-tuning. Our key idea is to view online selection not as static sample ranking, but as shaping the next target-oriented update under the optimizer state. We formulate this as an optimizer-aware update-matching problem, establish its connection to second-order target utility, and show why subset-level construction must account for interactions and redundancy among selected samples. Based on this view, we develop a two-stage Filter-then-Weight algorithm that first filters geometrically useful candidates and then optimizes their coefficients. To make the framework practical for LLMs, we introduce a factorized outer-product gradient representation and optimized matrix computations for long-context data. Experiments show that our method consistently improves convergence and downstream performance over existing online data selection baselines under the same data budget.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
MCMC-Correction of Score-Based Diffusion Models for Model Composition
arXiv:2307.14012v4 Announce Type: replace
Abstract: Diffusion models can be parameterized in terms of either score or energy function. The energy parameterization is attractive as it enables sampling procedures such as Markov Chain Monte Carlo (MCMC) that incorporates a Metropolis--Hastings (MH) correction step based on energy differences between proposed samples. Such corrections can significantly improve sampling quality, particularly in the context of model composition, where pre-trained models are combined to generate samples from novel distributions. Score-based diffusion models, on the other hand, are more widely adopted and come with a rich ecosystem of pre-trained models. However, they do not, in general, define an underlying energy function, making MH-based sampling inapplicable. In this work, we address this limitation by retaining score parameterization and introducing a novel MH-like acceptance rule based on line integration of the score function. This allows the reuse of existing diffusion models while still combining the reverse process with various MCMC techniques, viewed as an instance of annealed MCMC. Through experiments on synthetic and real-world data, we show that our MH-like samplers {yield relative improvements of similar magnitude to those observed} with energy-based models, without requiring explicit energy parameterization.
Fonte: arXiv stat.ML
Vision • Score 85
Improving Generalization of Deep Learning for Brain Metastases Segmentation Across Institutions
arXiv:2604.00397v1 Announce Type: new
Abstract: Background: Deep learning has demonstrated significant potential for automated brain metastases (BM) segmentation; however, models trained at a singular institution often exhibit suboptimal performance at various sites due to disparities in scanner hardware, imaging protocols, and patient demographics. The goal of this work is to create a domain adaptation framework that will allow for BM segmentation to be used across multiple institutions.
Methods: We propose a VAE-MMD preprocessing pipeline that combines variational autoencoders (VAE) with maximum mean discrepancy (MMD) loss, incorporating skip connections and self-attention mechanisms alongside nnU-Net segmentation. The method was tested on 740 patients from four public databases: Stanford, UCSF, UCLM, and PKG, evaluated by domain classifier's accuracy, sensitivity, precision, F1/F2 scores, surface Dice (sDice), and 95th percentile Hausdorff distance (HD95).
Results: VAE-MMD reduced domain classifier accuracy from 0.91 to 0.50, indicating successful feature alignment across institutions. Reconstructed volumes attained a PSNR greater than 36 dB, maintaining anatomical accuracy. The combined method raised the mean F1 by 11.1% (0.700 to 0.778), the mean sDice by 7.93% (0.7121 to 0.7686), and reduced the mean HD95 by 65.5% (11.33 to 3.91 mm) across all four centers compared to the baseline nnU-Net.
Conclusions: VAE-MMD effectively diminishes cross-institutional data heterogeneity and enhances BM segmentation generalization across volumetric, detection, and boundary-level metrics without necessitating target-domain labels, thereby overcoming a significant obstacle to the clinical implementation of AI-assisted segmentation.
Fonte: arXiv cs.CV
Privacy/Security/Fairness • Score 92
FecalFed: Privacy-Preserving Poultry Disease Detection via Federated Learning
arXiv:2604.00559v1 Announce Type: new
Abstract: Early detection of highly pathogenic avian influenza (HPAI) and endemic poultry diseases is critical for global food security. While computer vision models excel at classifying diseases from fecal imaging, deploying these systems at scale is bottlenecked by farm data privacy concerns and institutional data silos. Furthermore, existing open-source agricultural datasets frequently suffer from severe, undocumented data contamination. In this paper, we introduce $\textbf{FecalFed}$, a privacy-preserving federated learning framework for poultry disease classification. We first curate and release $\texttt{poultry-fecal-fl}$, a rigorously deduplicated dataset of 8,770 unique images across four disease classes, revealing and eliminating a 46.89$\%$ duplication rate in popular public repositories. To simulate realistic agricultural environments, we evaluate FecalFed under highly heterogeneous, non-IID conditions (Dirichlet $\alpha=0.5$). While isolated single-farm training collapses under this data heterogeneity, yielding only 64.86$\%$ accuracy, our federated approach recovers performance without centralizing sensitive data. Specifically, utilizing server-side adaptive optimization (FedAdam) with a Swin-Small architecture achieves 90.31$\%$ accuracy, closely approaching the centralized upper bound of 95.10\%. Furthermore, we demonstrate that an edge-optimized Swin-Tiny model maintains highly competitive performance at 89.74$\%$, establishing a highly efficient, privacy-first blueprint for on-farm avian disease monitoring.
Fonte: arXiv cs.CV
RL • Score 85
ShapE-GRPO: Shapley-Enhanced Reward Allocation for Multi-Candidate LLM Training
arXiv:2603.29871v1 Announce Type: new
Abstract: In user-agent interaction scenarios such as recommendation, brainstorming, and code suggestion, Large Language Models (LLMs) often generate sets of candidate recommendations where the objective is to maximize the collective utility of the entire set rather than individual candidates independently. However, existing reinforcement learning post-training paradigms, such as Group Relative Policy Optimization (GRPO), typically assign the same set-level scalar reward to every candidate in the set. This leads to noisy training signals where poor candidates free-ride on the high reward produced by a single strong peer, resulting in suboptimal exploration. To address this, we propose Shapley-Enhanced GRPO (ShapE-GRPO). By leveraging the permutation-invariant nature of set-level utility, we derive a Shapley-enhanced formulation from cooperative game theory to decompose set-level rewards into granular, candidate-specific signals. We show that our formulation preserves the fundamental axioms of the Shapley value while remaining computationally efficient with polynomial-time complexity. Empirically, ShapE-GRPO consistently outperforms standard GRPO across diverse datasets with accelerated convergence during training.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
C-TRAIL: A Commonsense World Framework for Trajectory Planning in Autonomous Driving
arXiv:2603.29908v1 Announce Type: new
Abstract: Trajectory planning for autonomous driving increasingly leverages large language models (LLMs) for commonsense reasoning, yet LLM outputs are inherently unreliable, posing risks in safety-critical applications. We propose C-TRAIL, a framework built on a Commonsense World that couples LLM-derived commonsense with a trust mechanism to guide trajectory planning. C-TRAIL operates through a closed-loop Recall, Plan, and Update cycle: the Recall module queries an LLM for semantic relations and quantifies their reliability via a dual-trust mechanism; the Plan module injects trust-weighted commonsense into Monte Carlo Tree Search (MCTS) through a Dirichlet trust policy; and the Update module adaptively refines trust scores and policy parameters from environmental feedback. Experiments on four simulated scenarios in Highway-env and two real-world levelXData datasets (highD, rounD) show that C-TRAIL consistently outperforms state-of-the-art baselines, reducing ADE by 40.2%, FDE by 51.7%, and improving SR by 16.9 percentage points on average. The source code is available at https://github.com/ZhihongCui/CTRAIL.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
CausalPulse: An Industrial-Grade Neurosymbolic Multi-Agent Copilot for Causal Diagnostics in Smart Manufacturing
arXiv:2603.29755v1 Announce Type: new
Abstract: Modern manufacturing environments demand real-time, trustworthy, and interpretable root-cause insights to sustain productivity and quality. Traditional analytics pipelines often treat anomaly detection, causal inference, and root-cause analysis as isolated stages, limiting scalability and explainability. In this work, we present CausalPulse, an industry-grade multi-agent copilot that automates causal diagnostics in smart manufacturing. It unifies anomaly detection, causal discovery, and reasoning through a neurosymbolic architecture built on standardized agentic protocols. CausalPulse is being deployed in a Robert Bosch manufacturing plant, integrating seamlessly with existing monitoring workflows and supporting real-time operation at production scale. Evaluations on both public (Future Factories) and proprietary (Planar Sensor Element) datasets show high reliability, achieving overall success rates of 98.0% and 98.73%. Per-criterion success rates reached 98.75% for planning and tool use, 97.3% for self-reflection, and 99.2% for collaboration. Runtime experiments report end-to-end latency of 50-60s per diagnostic workflow with near-linear scalability (R^2=0.97), confirming real-time readiness. Comparison with existing industrial copilots highlights distinct advantages in modularity, extensibility, and deployment maturity. These results demonstrate how CausalPulse's modular, human-in-the-loop design enables reliable, interpretable, and production-ready automation for next-generation manufacturing.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
OneComp: One-Line Revolution for Generative AI Model Compression
arXiv:2603.28845v1 Announce Type: new
Abstract: Deploying foundation models is increasingly constrained by memory footprint, latency, and hardware costs. Post-training compression can mitigate these bottlenecks by reducing the precision of model parameters without significantly degrading performance; however, its practical implementation remains challenging as practitioners navigate a fragmented landscape of quantization algorithms, precision budgets, data-driven calibration strategies, and hardware-dependent execution regimes. We present OneComp, an open-source compression framework that transforms this expert workflow into a reproducible, resource-adaptive pipeline. Given a model identifier and available hardware, OneComp automatically inspects the model, plans mixed-precision assignments, and executes progressive quantization stages, ranging from layer-wise compression to block-wise refinement and global refinement. A key architectural choice is treating the first quantized checkpoint as a deployable pivot, ensuring that each subsequent stage improves the same model and that quality increases as more compute is invested. By converting state-of-the-art compression research into an extensible, open-source, hardware-aware pipeline, OneComp bridges the gap between algorithmic innovation and production-grade model deployment.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
SimMOF: AI agent for Automated MOF Simulations
arXiv:2603.29152v1 Announce Type: new
Abstract: Metal-organic frameworks (MOFs) offer a vast design space, and as such, computational simulations play a critical role in predicting their structural and physicochemical properties. However, MOF simulations remain difficult to access because reliable analysis require expert decisions for workflow construction, parameter selection, tool interoperability, and the preparation of computational ready structures. Here, we introduce SimMOF, a large language model based multi agent framework that automates end-to-end MOF simulation workflows from natural language queries. SimMOF translates user requests into dependency aware plans, generates runnable inputs, orchestrates multiple agents to execute simulations, and summarizes results with analysis aligned to the user query. Through representative case studies, we show that SimMOF enables adaptive and cognitively autonomous workflows that reflect the iterative and decision driven behavior of human researchers and as such provides a scalable foundation for data driven MOF research.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Mimosa Framework: Toward Evolving Multi-Agent Systems for Scientific Research
arXiv:2603.28986v1 Announce Type: new
Abstract: Current Autonomous Scientific Research (ASR) systems, despite leveraging large language models (LLMs) and agentic architectures, remain constrained by fixed workflows and toolsets that prevent adaptation to evolving tasks and environments. We introduce Mimosa, an evolving multi-agent framework that automatically synthesizes task-specific multi-agent workflows and iteratively refines them through experimental feedback. Mimosa leverages the Model Context Protocol (MCP) for dynamic tool discovery, generates workflow topologies via a meta-orchestrator, executes subtasks through code-generating agents that invoke available tools and scientific software libraries, and scores executions with an LLM-based judge whose feedback drives workflow refinement. On ScienceAgentBench, Mimosa achieves a success rate of 43.1% with DeepSeek-V3.2, surpassing both single-agent baselines and static multi-agent configurations. Our results further reveal that models respond heterogeneously to multi-agent decomposition and iterative learning, indicating that the benefits of workflow evolution depend on the capabilities of the underlying execution model. Beyond these benchmarks, Mimosa modular architecture and tool-agnostic design make it readily extensible, and its fully logged execution traces and archived workflows support auditability by preserving every analytical step for inspection and potential replication. Combined with domain-expert guidance, the framework has the potential to automate a broad range of computationally accessible scientific tasks across disciplines. Released as a fully open-source platform, Mimosa aims to provide an open foundation for community-driven ASR.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Deep Learning-Based Anomaly Detection in Spacecraft Telemetry on Edge Devices
arXiv:2603.29375v1 Announce Type: new
Abstract: Spacecraft anomaly detection is critical for mission safety, yet deploying sophisticated models on-board presents significant challenges due to hardware constraints. This paper investigates three approaches for spacecraft telemetry anomaly detection -- forecasting & threshold, direct classification, and image classification -- and optimizes them for edge deployment using multi-objective neural architecture optimization on the European Space Agency Anomaly Dataset. Our baseline experiments demonstrate that forecasting & threshold achieves superior detection performance (92.7% Corrected Event-wise F0.5-score (CEF0.5)) [1] compared to alternatives. Through Pareto-optimal architecture optimization, we dramatically reduced computational requirements while maintaining capabilities -- the optimized forecasting & threshold model preserved 88.8% CEF0.5 while reducing RAM usage by 97.1% to just 59 KB and operations by 99.4%. Analysis of deployment viability shows our optimized models require just 0.36-6.25% of CubeSat RAM, making on-board anomaly detection practical even on highly constrained hardware. This research demonstrates that sophisticated anomaly detection capabilities can be successfully deployed within spacecraft edge computing constraints, providing near-instantaneous detection without exceeding hardware limitations or compromising mission safety.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Drift Estimation for Diffusion Processes Using Neural Networks Based on Discretely Observed Independent Paths
arXiv:2511.11161v2 Announce Type: replace
Abstract: This paper addresses the nonparametric estimation of the drift function over a compact domain for a time-homogeneous diffusion process, based on high-frequency discrete observations from $N$ independent trajectories. We propose a neural network-based estimator and derive a non-asymptotic convergence rate, decomposed into a training error, an approximation error, and a diffusion-related term scaling as ${\log N}/{N}$. For compositional drift functions, we establish an explicit rate. In the numerical experiments, we consider a drift function with local fluctuations generated by a double-layer compositional structure featuring local oscillations, and show that the empirical convergence rate becomes independent of the input dimension $d$. Compared to the $B$-spline method, the neural network estimator achieves better convergence rates and more effectively captures local features, particularly in higher-dimensional settings.
Fonte: arXiv stat.ML
MLOps/Systems • Score 85
Reasoning-Driven Synthetic Data Generation and Evaluation
arXiv:2603.29791v1 Announce Type: new
Abstract: Although many AI applications of interest require specialized multi-modal models, relevant data to train such models is inherently scarce or inaccessible. Filling these gaps with human annotators is prohibitively expensive, error-prone, and time-consuming, leading model builders to increasingly consider synthetic data as a scalable alternative. However, existing synthetic data generation methods often rely on manual prompts, evolutionary algorithms, or extensive seed data from the target distribution - limiting their scalability, explainability, and control. In this paper, we introduce Simula: a novel reasoning-driven framework for data generation and evaluation. It employs a seedless, agentic approach to generate synthetic datasets at scale, allowing users to define desired dataset characteristics through an explainable and controllable process that enables fine-grained resource allocation. We show the efficacy of our approach on a variety of datasets, rigorously testing both intrinsic and downstream properties. Our work (1) offers guidelines for synthetic data mechanism design, (2) provides insights into generating and evaluating synthetic data at scale, and (3) unlocks new opportunities for developing and deploying AI in domains where data scarcity or privacy concerns are paramount.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
Lie Generator Networks for Nonlinear Partial Differential Equations
arXiv:2603.29264v1 Announce Type: new
Abstract: Linear dynamical systems are fully characterized by their eigenspectra, accessible directly from the generator of the dynamics. For nonlinear systems governed by partial differential equations, no equivalent theory exists. We introduce Lie Generator Network--Koopman (LGN-KM), a neural operator that lifts nonlinear dynamics into a linear latent space and learns the continuous-time Koopman generator ($L_k$) through a decomposition $L_k = S - D_k$, where $S$ is skew-symmetric representing conservative inter-modal coupling, and $D_k$ is a positive-definite diagonal encoding modal dissipation. This architectural decomposition enforces stability and enables interpretability through direct spectral access to the learned dynamics. On two-dimensional Navier--Stokes turbulence, the generator recovers the known dissipation scaling and a complete multi-branch dispersion relation from trajectory data alone with no physics supervision. Independently trained models at different flow regimes recover matched gauge-invariant spectral structure, exposing a gauge freedom in the Koopman lifting. Because the generator is provably stable, it enables guaranteed long-horizon stability, continuous-time evaluation at arbitrary time, and physics-informed cross-viscosity model transfer.
Fonte: arXiv cs.LG
Evaluation/Benchmarks • Score 85
Spark-LLM-Eval: A Distributed Framework for Statistically Rigorous Large Language Model Evaluation
arXiv:2603.28769v1 Announce Type: cross
Abstract: Evaluating large language models at scale remains a practical bottleneck for many organizations. While existing evaluation frameworks work well for thousands of examples, they struggle when datasets grow to hundreds of thousands or millions of samples. This scale is common when assessing model behavior across diverse domains or conducting comprehensive regression testing. We present Spark-LLM-Eval, a distributed evaluation framework built natively on Apache Spark. The system treats evaluation as a data-parallel problem, partitioningexamplesacrossexecutorsandaggregatingresultswithproperstatistical accounting. Beyond raw throughput, we emphasize statistical rigor: every reported metric includes bootstrap confidence intervals, and model comparisons come with appropriate significance tests (paired t-tests, McNemar's test, or Wilcoxon signed-rank, depending on the metric type). The framework also addresses the cost problem inherent in LLM evaluation through content-addressable response caching backed by Delta Lake, which allows iterating on metric definitions without re-running inference. We describe the system architecture, the statistical methodology, and report benchmark results showing linear scaling with cluster size. The framework and all evaluation code are available as open source.
Fonte: arXiv cs.CL
MLOps/Systems • Score 85
Causality-inspired Federated Learning for Dynamic Spatio-Temporal Graphs
arXiv:2603.29384v1 Announce Type: new
Abstract: Federated Graph Learning (FGL) has emerged as a powerful paradigm for decentralized training of graph neural networks while preserving data privacy. However, existing FGL methods are predominantly designed for static graphs and rely on parameter averaging or distribution alignment, which implicitly assume that all features are equally transferable across clients, overlooking both the spatial and temporal heterogeneity and the presence of client-specific knowledge in real-world graphs. In this work, we identify that such assumptions create a vicious cycle of spurious representation entanglement, client-specific interference, and negative transfer, degrading generalization performance in Federated Learning over Dynamic Spatio-Temporal Graphs (FSTG). To address this issue, we propose a novel causality-inspired framework named SC-FSGL, which explicitly decouples transferable causal knowledge from client-specific noise through representation-level interventions. Specifically, we introduce a Conditional Separation Module that simulates soft interventions through client conditioned masks, enabling the disentanglement of invariant spatio-temporal causal factors from spurious signals and mitigating representation entanglement caused by client heterogeneity. In addition, we propose a Causal Codebook that clusters causal prototypes and aligns local representations via contrastive learning, promoting cross-client consistency and facilitating knowledge sharing across diverse spatio-temporal patterns. Experiments on five diverse heterogeneity Spatio-Temporal Graph (STG) datasets show that SC-FSGL outperforms state-of-the-art methods.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
From Physics to Surrogate Intelligence: A Unified Electro-Thermo-Optimization Framework for TSV Networks
arXiv:2603.29268v1 Announce Type: new
Abstract: High-density through-substrate vias (TSVs) enable 2.5D/3D heterogeneous integration but introduce significant signal-integrity and thermal-reliability challenges due to electrical coupling, insertion loss, and self-heating. Conventional full-wave finite-element method (FEM) simulations provide high accuracy but become computationally prohibitive for large design-space exploration. This work presents a scalable electro-thermal modeling and optimization framework that combines physics-informed analytical modeling, graph neural network (GNN) surrogates, and full-wave sign-off validation. A multi-conductor analytical model computes broadband S-parameters and effective anisotropic thermal conductivities of TSV arrays, achieving $5\%-10\%$ relative Frobenius error (RFE) across array sizes up to $15x15$. A physics-informed GNN surrogate (TSV-PhGNN), trained on analytical data and fine-tuned with HFSS simulations, generalizes to larger arrays with RFE below $2\%$ and nearly constant variance. The surrogate is integrated into a multi-objective Pareto optimization framework targeting reflection coefficient, insertion loss, worst-case crosstalk (NEXT/FEXT), and effective thermal conductivity. Millions of TSV configurations can be explored within minutes, enabling exhaustive layout and geometric optimization that would be infeasible using FEM alone. Final designs are validated with Ansys HFSS and Mechanical, showing strong agreement. The proposed framework enables rapid electro-thermal co-design of TSV arrays while reducing per-design evaluation time by more than six orders of magnitude.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
A Latent Risk-Aware Machine Learning Approach for Predicting Operational Success in Clinical Trials based on TrialsBank
arXiv:2603.29041v1 Announce Type: new
Abstract: Clinical trials are characterized by high costs, extended timelines, and substantial operational risk, yet reliable prospective methods for predicting trial success before initiation remain limited. Existing artificial intelligence approaches often focus on isolated metrics or specific development stages and frequently rely on variables unavailable at the trial design phase, limiting real-world applicability. We present a hierarchical latent risk-aware machine learning framework for prospective prediction of clinical trial operational success using a curated subset of TrialsBank, a proprietary AI-ready database developed by Sorintellis, comprising 13,700 trials. Operational success was defined as the ability to initiate, conduct, and complete a clinical trial according to planned timelines, recruitment targets, and protocol specifications through database lock. This approach decomposes operational success prediction into two modeling stages. First, intermediate latent operational risk factors are predicted using more than 180 drug- and trial-level features available before trial initiation. These predicted latent risks are then integrated into a downstream model to estimate the probability of operational success. A staged data-splitting strategy was employed to prevent information leakage, and models were benchmarked using XGBoost, CatBoost, and Explainable Boosting Machines. Across Phase I-III, the framework achieves strong out-of-sample performance, with F1-scores of 0.93, 0.92, and 0.91, respectively. Incorporating latent risk drivers improves discrimination of operational failures, and performance remains robust under independent inference evaluation. These results demonstrate that clinical trial operational success can be prospectively forecasted using a latent risk-aware AI framework, enabling early risk assessment and supporting data-driven clinical development decision-making.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
OptiMer: Optimal Distribution Vector Merging Is Better than Data Mixing for Continual Pre-Training
arXiv:2603.28858v1 Announce Type: new
Abstract: Continual pre-training is widely used to adapt LLMs to target languages and domains, yet the mixture ratio of training data remains a sensitive hyperparameter that is expensive to tune: they must be fixed before training begins, and a suboptimal choice can waste weeks of compute. In this work, we propose OptiMer, which decouples ratio selection from training: we train one CPT model per dataset, extract each model's distribution vector, which represents the parameter shift induced by that dataset, and search for optimal composition weights post-hoc via Bayesian optimization. Experiments on Gemma 3 27B across languages (Japanese, Chinese) and domains (Math, Code) show that OptiMer consistently outperforms data mixture and model averaging baselines with 15-35 times lower search cost. Key findings reveal that 1) the optimized weights can be interpreted as data mixture ratios, and retraining with these ratios improves data mixture CPT, and 2) the same vector pool can be re-optimized for a given objective without any retraining, producing target-tailored models on demand. Our work establishes that data mixture ratio selection, traditionally a pre-training decision, can be reformulated as a post-hoc optimization over distribution vectors, offering a more flexible paradigm for continual pre-training.
Fonte: arXiv cs.CL
Theory/Optimization • Score 85
DeepRV: Accelerating Spatiotemporal Inference with Pre-trained Neural Priors
arXiv:2503.21473v3 Announce Type: replace
Abstract: Gaussian Processes (GPs) provide a flexible and statistically principled foundation for modelling spatiotemporal phenomena, but their $O(N^3)$ scaling makes them intractable for large datasets. Approximate methods such as variational inference (VI), inducing-point (sparse) GPs, low-rank kernel approximations (e.g., Nystrom methods and random Fourier features), and approximations such as INLA improve scalability but typically trade off accuracy, calibration, or modelling flexibility. We introduce DeepRV, a neural-network surrogate that replaces GP prior sampling, while closely matching full GP accuracy at inference including hyperparameter estimates, and reducing computational complexity to $O(N^2)$, increasing scalability and inference speed. DeepRV serves as a drop-in replacement for GP prior realisations in e.g. MCMC-based probabilistic programming pipelines, preserving full model flexibility. Across simulated benchmarks, non-separable spatiotemporal GPs, and a real-world application to education deprivation in London (n = 4,994 locations), DeepRV achieves the highest fidelity to exact GPs while substantially accelerating inference. Code is provided in the dl4bi Python package, with all experiments run on a single consumer-grade GPU to ensure accessibility for practitioners.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Disentangled Graph Prompting for Out-Of-Distribution Detection
arXiv:2603.29644v1 Announce Type: new
Abstract: When testing data and training data come from different distributions, deep neural networks (DNNs) will face significant safety risks in practical applications. Therefore, out-of-distribution (OOD) detection techniques, which can identify OOD samples at test time and alert the system, are urgently needed. Existing graph OOD detection methods usually characterize fine-grained in-distribution (ID) patterns from multiple perspectives, and train end-to-end graph neural networks (GNNs) for prediction. However, due to the unavailability of OOD data during training, the absence of explicit supervision signals could lead to sub-optimal performance of end-to-end encoders. To address this issue, we follow the pre-training+prompting paradigm to utilize pre-trained GNN encoders, and propose Disentangled Graph Prompting (DGP), to capture fine-grained ID patterns with the help of ID graph labels. Specifically, we design two prompt generators that respectively generate class-specific and class-agnostic prompt graphs by modifying the edge weights of an input graph. We also design several effective losses to train the prompt generators and prevent trivial solutions. We conduct extensive experiments on ten datasets to demonstrate the superiority of our proposed DGP, which achieves a relative AUC improvement of 3.63% over the best graph OOD detection baseline. Ablation studies and hyper-parameter experiments further show the effectiveness of DGP. Code is available at https://github.com/BUPT-GAMMA/DGP.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Differentiable Initialization-Accelerated CPU-GPU Hybrid Combinatorial Scheduling
arXiv:2603.28943v1 Announce Type: new
Abstract: This paper presents a hybrid CPU-GPU framework for solving combinatorial scheduling problems formulated as Integer Linear Programming (ILP). While scheduling underpins many optimization tasks in computing systems, solving these problems optimally at scale remains a long-standing challenge due to their NP-hard nature. We introduce a novel approach that combines differentiable optimization with classical ILP solving. Specifically, we utilize differentiable presolving to rapidly generate high-quality partial solutions, which serve as warm-starts for commercial ILP solvers (CPLEX, Gurobi) and rising open-source solver HiGHS. This method enables significantly improved early pruning compared to state-of-the-art standalone solvers. Empirical results across industry-scale benchmarks demonstrate up to a $10\times$ performance gain over baselines, narrowing the optimality gap to $<0.1\%$. This work represents the first demonstration of utilizing differentiable optimization to initialize exact ILP solvers for combinatorial scheduling, opening new opportunities to integrate machine learning infrastructure with classical exact optimization methods across broader domains.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Derived Fields Preserve Fine-Scale Detail in Budgeted Neural Simulators
arXiv:2603.29224v1 Announce Type: new
Abstract: Fine-scale-faithful neural simulation under fixed storage budgets remains challenging. Many existing methods reduce high-frequency error by improving architectures, training objectives, or rollout strategies. However, under budgeted coarsen-quantize-decode pipelines, fine detail can already be lost when the carried state is constructed. In the canonical periodic incompressible Navier-Stokes setting, we show that primitive and derived fields undergo systematically different retained-band distortions under the same operator. Motivated by this observation, we formulate Derived-Field Optimization (DerivOpt), a general state-design framework that chooses which physical fields are carried and how storage budget is allocated across them under a calibrated channel model. Across the full time-dependent forward subset of PDEBench, DerivOpt not only improves pooled mean rollout nRMSE, but also delivers a decisive advantage in fine-scale fidelity over a broad set of strong baselines. More importantly, the gains are already visible at input time, before rollout learning begins. This indicates that the carried state is often the dominant bottleneck under tight storage budgets. These results suggest a broader conclusion: in budgeted neural simulation, carried-state design should be treated as a first-class design axis alongside architecture, loss, and rollout strategy.
Fonte: arXiv cs.LG
RL • Score 85
Realistic Market Impact Modeling for Reinforcement Learning Trading Environments
arXiv:2603.29086v1 Announce Type: new
Abstract: Reinforcement learning (RL) has shown promise for trading, yet most open-source backtesting environments assume negligible or fixed transaction costs, causing agents to learn trading behaviors that fail under realistic execution. We introduce three Gymnasium-compatible trading environments -- MACE (Market-Adjusted Cost Execution) stock trading, margin trading, and portfolio optimization -- that integrate nonlinear market impact models grounded in the Almgren-Chriss framework and the empirically validated square-root impact law. Each environment provides pluggable cost models, permanent impact tracking with exponential decay, and comprehensive trade-level logging. We evaluate five DRL algorithms (A2C, PPO, DDPG, SAC, TD3) on the NASDAQ-100, comparing a fixed 10 bps baseline against the AC model with Optuna-tuned hyperparameters. Our results show that (i) the cost model materially changes both absolute performance and the relative ranking of algorithms across all three environments; (ii) the AC model produces dramatically different trading behavior, e.g., daily costs dropping from $200k to $8k with turnover falling from 19% to 1%; (iii) hyperparameter optimization is essential for constraining pathological trading, with costs dropping up to 82%; and (iv) algorithm-cost model interactions are strongly environment-specific, e.g., DDPG's OOS Sharpe jumps from -2.1 to 0.3 under AC in margin trading while SAC's drops from -0.5 to -1.2. We release the full suite as an open-source extension to FinRL-Meta.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
PolarQuant: Optimal Gaussian Weight Quantization via Hadamard Rotation for LLM Compression
arXiv:2603.29078v1 Announce Type: new
Abstract: We present PolarQuant, a post-training weight quantization method for large language models (LLMs) that exploits the distributional structure of neural network weights to achieve near-lossless compression. PolarQuant operates in three stages: (1) block-wise normalization to the unit hypersphere, (2) Walsh-Hadamard rotation to transform coordinates into approximately Gaussian random variables, and (3) quantization with centroids matched to the Gaussian distribution. Our ablation reveals that Hadamard rotation alone accounts for 98% of the quality improvement, reducing Qwen3.5-9B perplexity from 6.90 (absmax Q5) to 6.40 (Delta = +0.03 from FP16), making it practically lossless without any calibration data. Furthermore, PolarQuant functions as an effective preprocessing step for downstream INT4 quantizers: PolarQuant Q5 dequantized and re-quantized by torchao INT4 achieves perplexity 6.56 versus 6.68 for direct absmax INT4, while maintaining 43.1 tok/s throughput at 6.5 GB VRAM. Code and models are publicly available.
Fonte: arXiv cs.CL
Applications • Score 85
Owl-AuraID 1.0: An Intelligent System for Autonomous Scientific Instrumentation and Scientific Data Analysis
arXiv:2603.29828v1 Announce Type: new
Abstract: Scientific discovery increasingly depends on high-throughput characterization, yet automation is hindered by proprietary GUIs and the limited generalizability of existing API-based systems. We present Owl-AuraID, a software-hardware collaborative embodied agent system that adopts a GUI-native paradigm to operate instruments through the same interfaces as human experts. Its skill-centric framework integrates Type-1 (GUI operation) and Type-2 (data analysis) skills into end-to-end workflows, connecting physical sample handling with scientific interpretation. Owl-AuraID demonstrates broad coverage across ten categories of precision instruments and diverse workflows, including multimodal spectral analysis, microscopic imaging, and crystallographic analysis, supporting modalities such as FTIR, NMR, AFM, and TGA. Overall, Owl-AuraID provides a practical, extensible foundation for autonomous laboratories and illustrates a path toward evolving laboratory intelligence through reusable operational and analytical skills. The code are available at https://github.com/OpenOwlab/AuraID.
Fonte: arXiv cs.AI
MLOps/Systems • Score 88
ARCS: Autoregressive Circuit Synthesis with Topology-Aware Graph Attention and Spec Conditioning
arXiv:2603.29068v1 Announce Type: new
Abstract: I present ARCS, a system for amortized analog circuit generation that produces complete, SPICE-simulatable designs (topology and component values) in milliseconds rather than the minutes required by search-based methods. A hybrid pipeline combining two learned generators (a graph VAE and a flow-matching model) with SPICE-based ranking achieves 99.9% simulation validity (reward 6.43/8.0) across 32 topologies using only 8 SPICE evaluations, 40x fewer than genetic algorithms. For single-model inference, a topology-aware Graph Transformer with Best-of-3 candidate selection reaches 85% simulation validity in 97ms, over 600x faster than random search. The key technical contribution is Group Relative Policy Optimization (GRPO): I identify a critical failure mode of REINFORCE (cross-topology reward distribution mismatch) and resolve it with per-topology advantage normalization, improving simulation validity by +9.6pp over REINFORCE in only 500 RL steps (10x fewer). Grammar-constrained decoding additionally guarantees 100% structural validity by construction via topology-aware token masking. ARCS does not yet match the per-design quality of search-based optimization (5.48 vs. 7.48 reward), but its >1000x speed advantage enables rapid prototyping, design-space exploration, and warm-starting search methods (recovering 96.6% of GA quality with 49% fewer simulations).
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
mtslearn: Machine Learning in Python for Medical Time Series
arXiv:2603.29432v1 Announce Type: new
Abstract: Medical time-series data captures the dynamic progression of patient conditions, playing a vital role in modern clinical decision support systems. However, real-world clinical data is highly heterogeneous and inconsistently formatted. Furthermore, existing machine learning tools often have steep learning curves and fragmented workflows. Consequently, a significant gap remains between cutting-edge AI technologies and clinical application. To address this, we introduce mtslearn, an end-to-end integrated toolkit specifically designed for medical time-series data. First, the framework provides a unified data interface that automates the parsing and alignment of wide, long, and flat data formats. This design significantly reduces data cleaning overhead. Building on this, mtslearn provides a complete pipeline from data reading and feature engineering to model training and result visualization. Furthermore, it offers flexible interfaces for custom algorithms. Through a modular design, mtslearn simplifies complex data engineering tasks into a few lines of code. This significantly lowers the barrier to entry for clinicians with limited programming experience, empowering them to focus more on exploring medical hypotheses and accelerating the translation of advanced algorithms into real-world clinical practice. mtslearn is publicly available at https://github.com/PKUDigitalHealth/mtslearn.
Fonte: arXiv cs.LG
Vision • Score 85
Dual-Imbalance Continual Learning for Real-World Food Recognition
arXiv:2603.29133v1 Announce Type: new
Abstract: Visual food recognition in real-world dietary logging scenarios naturally exhibits severe data imbalance, where a small number of food categories appear frequently while many others occur rarely, resulting in long-tailed class distributions. In practice, food recognition systems often operate in a continual learning setting, where new categories are introduced sequentially over time. However, existing studies typically assume that each incremental step introduces a similar number of new food classes, which rarely happens in real world where the number of newly observed categories can vary significantly across steps, leading to highly uneven learning dynamics. As a result, continual food recognition exhibits a dual imbalance: imbalanced samples within each food class and imbalanced numbers of new food classes to learn at each incremental learning step. In this work, we introduce DIME, a Dual-Imbalance-aware Adapter Merging framework for continual food recognition. DIME learns lightweight adapters for each task using parameter-efficient fine-tuning and progressively integrates them through a class-count guided spectral merging strategy. A rank-wise threshold modulation mechanism further stabilizes the merging process by preserving dominant knowledge while allowing adaptive updates. The resulting model maintains a single merged adapter for inference, enabling efficient deployment without accumulating task-specific modules. Experiments on realistic long-tailed food benchmarks under our step-imbalanced setup show that the proposed method consistently improves by more than 3% over the strongest existing continual learning baselines. Code is available at https://github.com/xiaoyanzhang1/DIME.
Fonte: arXiv cs.CV
Vision • Score 85
Fisheye3R: Adapting Unified 3D Feed-Forward Foundation Models to Fisheye Lenses
arXiv:2603.28896v1 Announce Type: new
Abstract: Feed-forward foundation models for multi-view 3-dimensional (3D) reconstruction have been trained on large-scale datasets of perspective images; when tested on wide field-of-view images, e.g., from a fisheye camera, their performance degrades. Their error arises from changes in spatial positions of pixels due to a non-linear projection model that maps 3D points onto the 2D image plane. While one may surmise that training on fisheye images would resolve this problem, there are far fewer fisheye images with ground truth than perspective images, which limit generalization. To enable inference on imagery exhibiting high radial distortion, we propose Fisheye3R, a novel adaptation framework that extends these multi-view 3D reconstruction foundation models to natively accommodate fisheye inputs without performance regression on perspective images. To address the scarcity of fisheye images and ground truth, we introduce flexible learning schemes that support self-supervised adaptation using only unlabeled perspective images and supervised adaptation without any fisheye training data. Extensive experiments across three foundation models, including VGGT, $\pi^3$, and MapAnything, demonstrate that our approach consistently improves camera pose, depth, point map, and field-of-view estimation on fisheye images.
Fonte: arXiv cs.CV
Vision • Score 85
M2H-MX: Multi-Task Dense Visual Perception for Real-Time Monocular Spatial Understanding
arXiv:2603.29236v1 Announce Type: new
Abstract: Monocular cameras are attractive for robotic perception due to their low cost and ease of deployment, yet achieving reliable real-time spatial understanding from a single image stream remains challenging. While recent multi-task dense prediction models have improved per-pixel depth and semantic estimation, translating these advances into stable monocular mapping systems is still non-trivial.
This paper presents M2H-MX, a real-time multi-task perception model for monocular spatial understanding. The model preserves multi-scale feature representations while introducing register-gated global context and controlled cross-task interaction in a lightweight decoder, enabling depth and semantic predictions to reinforce each other under strict latency constraints. Its outputs integrate directly into an unmodified monocular SLAM pipeline through a compact perception-to-mapping interface.
We evaluate both dense prediction accuracy and in-the-loop system performance. On NYUDv2, M2H-MX-L achieves state-of-the-art results, improving semantic mIoU by 6.6% and reducing depth RMSE by 9.4% over representative multi-task baselines. When deployed in a real-time monocular mapping system on ScanNet, M2H-MX reduces average trajectory error by 60.7% compared to a strong monocular SLAM baseline while producing cleaner metric-semantic maps. These results demonstrate that modern multi-task dense prediction can be reliably deployed for real-time monocular spatial perception in robotic systems.
Fonte: arXiv cs.CV
Theory/Optimization • Score 85
mlr3mbo: Bayesian Optimization in R
arXiv:2603.29730v1 Announce Type: new
Abstract: We present mlr3mbo, a comprehensive and modular toolbox for Bayesian optimization in R. mlr3mbo supports single- and multi-objective optimization, multi-point proposals, batch and asynchronous parallelization, input and output transformations, and robust error handling. While it can be used for many standard Bayesian optimization variants in applied settings, researchers can also construct custom BO algorithms from its flexible building blocks. In addition to an introduction to the software, its design principles, and its building blocks, the paper presents two extensive empirical evaluations of the software on the surrogate-based benchmark suite YAHPO Gym. To identify robust default configurations for both numeric and mixed-hierarchical optimization regimes, and to gain further insights into the respective impacts of individual settings, we run a coordinate descent search over the mlr3mbo configuration space and analyze its results. Furthermore, we demonstrate that mlr3mbo achieves state-of-the-art performance by benchmarking it against a wide range of optimizers, including HEBO, SMAC3, Ax, and Optuna.
Fonte: arXiv stat.ML
RL • Score 85
Enhancing Policy Learning with World-Action Model
arXiv:2603.28955v1 Announce Type: new
Abstract: This paper presents the World-Action Model (WAM), an action-regularized world model that jointly reasons over future visual observations and the actions that drive state transitions. Unlike conventional world models trained solely via image prediction, WAM incorporates an inverse dynamics objective into DreamerV2 that predicts actions from latent state transitions, encouraging the learned representations to capture action-relevant structure critical for downstream control. We evaluate WAM on enhancing policy learning across eight manipulation tasks from the CALVIN benchmark. We first pretrain a diffusion policy via behavioral cloning on world model latents, then refine it with model-based PPO inside the frozen world model. Without modifying the policy architecture or training procedure, WAM improves average behavioral cloning success from 59.4% to 71.2% over DreamerV2 and DiWA baselines. After PPO fine-tuning, WAM achieves 92.8% average success versus 79.8% for the baseline, with two tasks reaching 100%, using 8.7x fewer training steps.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
Epistemic Errors of Imperfect Multitask Learners When Distributions Shift
arXiv:2505.23496v3 Announce Type: replace-cross
Abstract: Uncertainty-aware machine learners, such as Bayesian neural networks, output a quantification of uncertainty instead of a point prediction. We provide uncertainty-aware learners with a principled framework to characterize, and identify ways to eliminate, errors that arise from reducible (epistemic) uncertainty. We introduce a principled definition of epistemic error, and provide a decompositional epistemic error bound which operates in the very general setting of imperfect multitask learning under distribution shift. In this setting, the training (source) data may arise from multiple tasks, the test (target) data may differ systematically from the source data tasks, and/or the learner may not arrive at an accurate characterization of the source data. Our bound separately attributes epistemic errors to each of multiple aspects of the learning procedure and environment. As corollaries of the general result, we provide epistemic error bounds specialized to the settings of Bayesian transfer learning and distribution shift within $\epsilon$-neighborhoods.
Fonte: arXiv stat.ML
Theory/Optimization • Score 85
Expressive Power of Implicit Models: Rich Equilibria and Test-Time Scaling
arXiv:2510.03638v4 Announce Type: replace-cross
Abstract: Implicit models, an emerging model class, compute outputs by iterating a single parameter block to a fixed point. This architecture realizes an infinite-depth, weight-tied network that trains with constant memory, significantly reducing memory needs for the same level of performance compared to explicit models. While it is empirically known that these compact models can often match or even exceed the accuracy of larger explicit networks by allocating more test-time compute, the underlying mechanism remains poorly understood.
We study this gap through a nonparametric analysis of expressive power. We provide a strict mathematical characterization, showing that a simple and regular implicit operator can, through iteration, progressively express more complex mappings. We prove that for a broad class of implicit models, this process lets the model's expressive power scale with test-time compute, ultimately matching a much richer function class. The theory is validated across four domains: image reconstruction, scientific computing, operations research, and LLM reasoning, demonstrating that as test-time iterations increase, the complexity of the learned mapping rises, while the solution quality simultaneously improves and stabilizes.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
MemFactory: Unified Inference & Training Framework for Agent Memory
arXiv:2603.29493v1 Announce Type: new
Abstract: Memory-augmented Large Language Models (LLMs) are essential for developing capable, long-term AI agents. Recently, applying Reinforcement Learning (RL) to optimize memory operations, such as extraction, updating, and retrieval, has emerged as a highly promising research direction. However, existing implementations remain highly fragmented and task-specific, lacking a unified infrastructure to streamline the integration, training, and evaluation of these complex pipelines. To address this gap, we present MemFactory, the first unified, highly modular training and inference framework specifically designed for memory-augmented agents. Inspired by the success of unified fine-tuning frameworks like LLaMA-Factory, MemFactory abstracts the memory lifecycle into atomic, plug-and-play components, enabling researchers to seamlessly construct custom memory agents via a "Lego-like" architecture. Furthermore, the framework natively integrates Group Relative Policy Optimization (GRPO) to fine-tune internal memory management policies driven by multi-dimensional environmental rewards. MemFactory provides out-of-the-box support for recent cutting-edge paradigms, including Memory-R1, RMM, and MemAgent. We empirically validate MemFactory on the open-source MemAgent architecture using its publicly available training and evaluation data. Across both in-domain and out-of-distribution evaluation sets, MemFactory consistently improves performance over the corresponding base models, with relative gains of up to 14.8%. By providing a standardized, extensible, and easy-to-use infrastructure, MemFactory significantly lowers the barrier to entry, paving the way for future innovations in memory-driven AI agents.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
Calibrated Confidence Expression for Radiology Report Generation
arXiv:2603.29492v1 Announce Type: new
Abstract: Safe deployment of Large Vision-Language Models (LVLMs) in radiology report generation requires not only accurate predictions but also clinically interpretable indicators of when outputs should be thoroughly reviewed, enabling selective radiologist verification and reducing the risk of hallucinated findings influencing clinical decisions. One intuitive approach to this is verbalized confidence, where the model explicitly states its certainty. However, current state-of-the-art language models are often overconfident, and research on calibration in multimodal settings such as radiology report generation is limited. To address this gap, we introduce ConRad (Confidence Calibration for Radiology Reports), a reinforcement learning framework for fine-tuning medical LVLMs to produce calibrated verbalized confidence estimates alongside radiology reports. We study two settings: a single report-level confidence score and a sentence-level variant assigning a confidence to each claim. Both are trained using the GRPO algorithm with reward functions based on the logarithmic scoring rule, which incentivizes truthful self-assessment by penalizing miscalibration and guarantees optimal calibration under reward maximization. Experimentally, ConRad substantially improves calibration and outperforms competing methods. In a clinical evaluation we show that ConRad's report level scores are well aligned with clinicians' judgment. By highlighting full reports or low-confidence statements for targeted review, ConRad can support safer clinical integration of AI-assistance for report generation.
Fonte: arXiv cs.CL
NLP/LLMs • Score 85
Developing Adaptive Context Compression Techniques for Large Language Models (LLMs) in Long-Running Interactions
arXiv:2603.29193v1 Announce Type: new
Abstract: Large Language Models (LLMs) often experience performance degradation during long-running interactions due to increasing context length, memory saturation, and computational overhead. This paper presents an adaptive context compression framework that integrates importance-aware memory selection, coherence-sensitive filtering, and dynamic budget allocation to retain essential conversational information while controlling context growth. The approach is evaluated on LOCOMO, LOCCO, and LongBench benchmarks to assess answer quality, retrieval accuracy, coherence preservation, and efficiency. Experimental results demonstrate that the proposed method achieves consistent improvements in conversational stability and retrieval performance while reducing token usage and inference latency compared with existing memory and compression-based approaches. These findings indicate that adaptive context compression provides an effective balance between long-term memory preservation and computational efficiency in persistent LLM interactions
Fonte: arXiv cs.CV
NLP/LLMs • Score 85
Kwame 2.0: Human-in-the-Loop Generative AI Teaching Assistant for Large Scale Online Coding Education in Africa
arXiv:2603.29159v1 Announce Type: new
Abstract: Providing timely and accurate learning support in large-scale online coding courses is challenging, particularly in resource-constrained contexts. We present Kwame 2.0, a bilingual (English-French) generative AI teaching assistant built using retrieval-augmented generation and deployed in a human-in-the-loop forum within SuaCode, an introductory mobile-based coding course for learners across Africa. Kwame 2.0 retrieves relevant course materials and generates context-aware responses while encouraging human oversight and community participation. We deployed the system in a 15-month longitudinal study spanning 15 cohorts with 3,717 enrollments across 35 African countries. Evaluation using community feedback and expert ratings shows that Kwame 2.0 provided high-quality and timely support, achieving high accuracy on curriculum-related questions, while human facilitators and peers effectively mitigated errors, particularly for administrative queries. Our findings demonstrate that human-in-the-loop generative AI systems can combine the scalability and speed of AI with the reliability of human support, offering an effective approach to learning assistance for underrepresented populations in resource-constrained settings at scale.
Fonte: arXiv cs.CL
RL • Score 85
APEX-EM: Non-Parametric Online Learning for Autonomous Agents via Structured Procedural-Episodic Experience Replay
arXiv:2603.29093v1 Announce Type: new
Abstract: LLM-based autonomous agents lack persistent procedural memory: they re-derive solutions from scratch even when structurally identical tasks have been solved before. We present \textbf{APEX-EM}, a non-parametric online learning framework that accumulates, retrieves, and reuses structured procedural plans without modifying model weights. APEX-EM introduces: (1) a \emph{structured experience representation} encoding the full procedural-episodic trace of each execution -- planning steps, artifacts, iteration history with error analysis, and quality scores; (2) a \emph{Plan-Retrieve-Generate-Iterate-Ingest} (PRGII) workflow with Task Verifiers providing multi-dimensional reward signals; and (3) a \emph{dual-outcome Experience Memory} with hybrid retrieval combining semantic search, structural signature matching, and plan DAG traversal -- enabling cross-domain transfer between tasks sharing no lexical overlap but analogous operational structure. Successful experiences serve as positive in-context examples; failures as negative examples with structured error annotations.
We evaluate on BigCodeBench~\cite{zhuo2025bigcodebench}, KGQAGen-10k~\cite{zhang2025kgqagen}, and Humanity's Last Exam~\cite{phan2025hle} using Claude Sonnet 4.5 and Opus 4.5. On KGQAGen-10k, APEX-EM achieves 89.6\% accuracy versus 41.3\% without memory (+48.3pp), surpassing the oracle-retrieval upper bound (84.9\%). On BigCodeBench, it reaches 83.3\% SR from a 53.9\% baseline (+29.4pp), exceeding MemRL's~\cite{memrl2025} +11.0pp gain under comparable frozen-backbone conditions (noting backbone differences controlled for in our analysis). On HLE, entity graph retrieval reaches 48.0\% from 25.2\% (+22.8pp). Ablations show component value is task-dependent: rich judge feedback is negligible for code generation but critical for structured queries (+10.3pp), while binary-signal iteration partially compensates for weaker feedback.
Fonte: arXiv cs.CL
Theory/Optimization • Score 85
Foundations of Polar Linear Algebra
arXiv:2603.28939v1 Announce Type: new
Abstract: This work revisits operator learning from a spectral perspective by introducing Polar Linear Algebra, a structured framework based on polar geometry that combines a linear radial component with a periodic angular component. Starting from this formulation, we define the associated operators and analyze their spectral properties. As a proof of feasibility, the framework is evaluated on a canonical benchmark (MNIST). Despite the simplicity of the task, the results demonstrate that polar and fully spectral operators can be trained reliably, and that imposing self-adjoint-inspired spectral constraints improves stability and convergence. Beyond accuracy, the proposed formulation leads to a reduction in parameter count and computational complexity, while providing a more interpretable representation in terms of decoupled spectral modes. By moving from a spatial to a spectral domain, the problem decomposes into orthogonal eigenmodes that can be treated as independent computational pipelines. This structure naturally exposes an additional dimension of model parallelization, complementing existing parallel strategies without relying on ad-hoc partitioning. Overall, the work offers a different conceptual lens for operator learning, particularly suited to problems where spectral structure and parallel execution are central.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
On computing and the complexity of computing higher-order $U$-statistics, exactly
arXiv:2508.12627v2 Announce Type: replace
Abstract: Higher-order $U$-statistics abound in fields such as statistics, machine learning, and computer science, but are known to be highly time-consuming to compute in practice. Despite their widespread appearance, a comprehensive study of their computational complexity is surprisingly lacking. This paper aims to fill this gap by presenting several results related to the computational aspect of $U$-statistics. First, we derive a useful decomposition from a $m$-th order $U$-statistic to a linear combination of $V$-statistics with orders not exceeding $m$, which are generally more feasible to compute. Second, we explore the connection between exactly computing $V$-statistics and Einstein summation, a tool often used in computational mathematics and quantum computing to accelerate tensor computations. Third, we provide an optimistic estimate of the time complexity for exactly computing $U$-statistics, based on the treewidth of a particular graph associated with the $U$-statistic kernel. The above ingredients lead to (1) a new, much more runtime-efficient algorithm to exactly compute general higher-order $U$-statistics, and (2) a more streamlined characterization of runtime complexity of computing $U$-statistics. We develop an accompanying open-source package called \texttt{u-stats} in both Python (https://github.com/zrq1706/U-Statistics-Python) and R (https://github.com/cxy0714/U-Statistics-R). We demonstrate through three examples in statistics that \texttt{u-stats} achieves impressive runtime performance compared to existing benchmarks. This paper also aspires to achieve two goals: (1) to capture the interest of researchers in both statistics and other related areas to further advance the algorithmic development of $U$-statistics and (2) to lift the burden of implementing higher-order $U$-statistics from practitioners.
Fonte: arXiv stat.ML
RecSys • Score 85
Monodense Deep Neural Model for Determining Item Price Elasticity
arXiv:2603.29261v1 Announce Type: new
Abstract: Item Price Elasticity is used to quantify the responsiveness of consumer demand to changes in item prices, enabling businesses to create pricing strategies and optimize revenue management. Sectors such as store retail, e-commerce, and consumer goods rely on elasticity information derived from historical sales and pricing data. This elasticity provides an understanding of purchasing behavior across different items, consumer discount sensitivity, and demand elastic departments. This information is particularly valuable for competitive markets and resource-constrained businesses decision making which aims to maximize profitability and market share. Price elasticity also uncovers historical shifts in consumer responsiveness over time. In this paper, we model item-level price elasticity using large-scale transactional datasets, by proposing a novel elasticity estimation framework which has the capability to work in an absence of treatment control setting. We test this framework by using Machine learning based algorithms listed below, including our newly proposed Monodense deep neural network.
(1) Monodense-DL network -- Hybrid neural network architecture combining embedding, dense, and Monodense layers (2) DML -- Double machine learning setting using regression models (3) LGBM -- Light Gradient Boosting Model
We evaluate our model on multi-category retail data spanning millions of transactions using a back testing framework. Experimental results demonstrate the superiority of our proposed neural network model within the framework compared to other prevalent ML based methods listed above.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Improving Efficiency of GPU Kernel Optimization Agents using a Domain-Specific Language and Speed-of-Light Guidance
arXiv:2603.29010v1 Announce Type: new
Abstract: Optimizing GPU kernels with LLM agents is an iterative process over a large design space. Every candidate must be generated, compiled, validated, and profiled, so fewer trials will save both runtime and cost. We make two key observations. First, the abstraction level that agents operate at is important. If it is too low, the LLM wastes reasoning on low-impact details. If it is too high, it may miss important optimization choices. Second, agents cannot easily tell when they reach the point of diminishing returns, wasting resources as they continue searching.
These observations motivate two design principles to improve efficiency: (1) a compact domain-specific language (DSL) that can be learned in context and lets the model reason at a higher level while preserving important optimization levers, and (2) Speed-of-Light (SOL) guidance that uses first-principles performance bounds to steer and budget search. We implement these principles in $\mu$CUTLASS, a DSL with a compiler for CUTLASS-backed GPU kernels that covers kernel configuration, epilogue fusion, and multi-stage pipelines. We use SOL guidance to estimate headroom and guide optimization trials, deprioritize problems that are near SOL, and flag kernels that game the benchmark.
On 59 KernelBench problems with the same iteration budgets, switching from generating low-level code to DSL code using GPT-5-mini turns a 0.40x geomean regression into a 1.27x speedup over PyTorch. Adding SOL-guided steering raises this to 1.56x. Across model tiers, $\mu$CUTLASS + SOL-guidance lets weaker models outperform stronger baseline agents at lower token cost. SOL-guided budgeting saves 19-43% of tokens while retaining at least 95% of geomean speedup, with the best policy reaching a 1.68x efficiency gain. Lastly, SOL analysis helps detect benchmark-gaming cases, where kernels may appear fast while failing to perform the intended computation.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Stochastic Dimension Implicit Functional Projections for Exact Integral Conservation in High-Dimensional PINNs
arXiv:2603.29237v1 Announce Type: new
Abstract: Enforcing exact macroscopic conservation laws, such as mass and energy, in neural partial differential equation (PDE) solvers is computationally challenging in high dimensions. Traditional discrete projections rely on deterministic quadrature that scales poorly and restricts mesh-free formulations like PINNs. Furthermore, high-order operators incur heavy memory overhead, and generic optimization often lacks convergence guarantees for non-convex conservation manifolds. To address this, we propose the Stochastic Dimension Implicit Functional Projection (SDIFP) framework. Instead of projecting discrete vectors, SDIFP applies a global affine transformation to the continuous network output. This yields closed-form solutions for integral constraints via detached Monte Carlo (MC) quadrature, bypassing spatial grid dependencies. For scalable training, we introduce a doubly-stochastic unbiased gradient estimator (DS-UGE). By decoupling spatial sampling from differential operator subsampling, the DS-UGE reduces memory complexity from $\mathcal{O}(M \times N_{\mathcal{L}})$ to $\mathcal{O}(N \times |\mathcal{I}|)$. SDIFP mitigates sampling variance, preserves solution regularity, and maintains $\mathcal{O}(1)$ inference efficiency, providing a scalable, mesh-free approach for solving conservative high-dimensional PDEs.
Fonte: arXiv cs.LG
NLP/LLMs • Score 85
Quality-Controlled Active Learning via Gaussian Processes for Robust Structure-Property Learning in Autonomous Microscopy
arXiv:2603.29135v1 Announce Type: new
Abstract: Autonomous experimental systems are increasingly used in materials research to accelerate scientific discovery, but their performance is often limited by low-quality, noisy data. This issue is especially problematic in data-intensive structure-property learning tasks such as Image-to-Spectrum (Im2Spec) and Spectrum-to-Image (Spec2Im) translations, where standard active learning strategies can mistakenly prioritize poor-quality measurements. We introduce a gated active learning framework that combines curiosity-driven sampling with a physics-informed quality control filter based on the Simple Harmonic Oscillator model fits, allowing the system to automatically exclude low-fidelity data during acquisition. Evaluations on a pre-acquired dataset of band-excitation piezoresponse spectroscopy (BEPS) data from PbTiO3 thin films with spatially localized noise show that the proposed method outperforms random sampling, standard active learning, and multitask learning strategies. The gated approach enhances both Im2Spec and Spec2Im by handling noise during training and acquisition, leading to more reliable forward and inverse predictions. In contrast, standard active learners often misinterpret noise as uncertainty and end up acquiring bad samples that hurt performance. Given its promising applicability, we further deployed the framework in real-time experiments on BiFeO3 thin films, demonstrating its effectiveness in real autonomous microscopy experiments. Overall, this work supports a shift toward hybrid autonomy in self-driving labs, where physics-informed quality assessment and active decision-making work hand-in-hand for more reliable discovery.
Fonte: arXiv cs.LG
NLP/LLMs • Score 88
AgentFixer: From Failure Detection to Fix Recommendations in LLM Agentic Systems
arXiv:2603.29848v1 Announce Type: new
Abstract: We introduce a comprehensive validation framework for LLM-based agentic systems that provides systematic diagnosis and improvement of reliability failures. The framework includes fifteen failure-detection tools and two root-cause analysis modules that jointly uncover weaknesses across input handling, prompt design, and output generation. It integrates lightweight rule-based checks with LLM-as-a-judge assessments to support structured incident detection, classification, and repair. We applied the framework to IBM CUGA, evaluating its performance on the AppWorld and WebArena benchmarks. The analysis revealed recurrent planner misalignments, schema violations, brittle prompt dependencies, and more. Based on these insights, we refined both prompting and coding strategies, maintaining CUGA's benchmark results while enabling mid-sized models such as Llama 4 and Mistral Medium to achieve notable accuracy gains, substantially narrowing the gap with frontier models. Beyond quantitative validation, we conducted an exploratory study that fed the framework's diagnostic outputs and agent description into an LLM for self-reflection and prioritization. This interactive analysis produced actionable insights on recurring failure patterns and focus areas for improvement, demonstrating how validation itself can evolve into an agentic, dialogue-driven process. These results show a path toward scalable, quality assurance, and adaptive validation in production agentic systems, offering a foundation for more robust, interpretable, and self-improving agentic architectures.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
View-oriented Conversation Compiler for Agent Trace Analysis
arXiv:2603.29678v1 Announce Type: new
Abstract: Agent traces carry increasing analytical value in the era of context learning and harness-driven agentic cognition, yet most prior work treats conversation format as a trivial engineering detail. Modern agent conversations contain deeply structured content, including nested tool calls and results, chain-of-thought reasoning blocks, sub-agent invocations, context-window compaction boundaries, and harness-injected system directives, whose complexity far exceeds that of simple user-assistant exchanges. Feeding such traces to a reflector or other analytical mechanism in plain text, JSON, YAML, or via grep can materially degrade analysis quality. This paper presents VCC (View-oriented Conversation Compiler), a compiler (lex, parse, IR, lower, emit) that transforms raw agent JSONL logs into a family of structured views: a full view (lossless transcript serving as the canonical line-number coordinate system), a user-interface view (reconstructing the interaction as the user actually perceived it), and an adaptive view (a structure-preserving projection governed by a relevance predicate). In a context-learning experiment on AppWorld, replacing only the reflector's input format, from raw JSONL to VCC-compiled views, leads to higher pass rates across all three model configurations tested, while cutting reflector token consumption by half to two-thirds and producing more concise learned memory. These results suggest that message format functions as infrastructure for context learning, not as an incidental implementation choice.
Fonte: arXiv cs.AI
MLOps/Systems • Score 85
ELT-Bench-Verified: Benchmark Quality Issues Underestimate AI Agent Capabilities
arXiv:2603.29399v1 Announce Type: new
Abstract: Constructing Extract-Load-Transform (ELT) pipelines is a labor-intensive data engineering task and a high-impact target for AI automation. On ELT-Bench, the first benchmark for end-to-end ELT pipeline construction, AI agents initially showed low success rates, suggesting they lacked practical utility.
We revisit these results and identify two factors causing a substantial underestimation of agent capabilities. First, re-evaluating ELT-Bench with upgraded large language models reveals that the extraction and loading stage is largely solved, while transformation performance improves significantly. Second, we develop an Auditor-Corrector methodology that combines scalable LLM-driven root-cause analysis with rigorous human validation (inter-annotator agreement Fleiss' kappa = 0.85) to audit benchmark quality. Applying this to ELT-Bench uncovers that most failed transformation tasks contain benchmark-attributable errors -- including rigid evaluation scripts, ambiguous specifications, and incorrect ground truth -- that penalize correct agent outputs.
Based on these findings, we construct ELT-Bench-Verified, a revised benchmark with refined evaluation logic and corrected ground truth. Re-evaluating on this version yields significant improvement attributable entirely to benchmark correction. Our results show that both rapid model improvement and benchmark quality issues contributed to underestimating agent capabilities. More broadly, our findings echo observations of pervasive annotation errors in text-to-SQL benchmarks, suggesting quality issues are systemic in data engineering evaluation. Systematic quality auditing should be standard practice for complex agentic tasks. We release ELT-Bench-Verified to provide a more reliable foundation for progress in AI-driven data engineering automation.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Liquid Networks with Mixture Density Heads for Efficient Imitation Learning
arXiv:2603.27058v1 Announce Type: new
Abstract: We compare liquid neural networks with mixture density heads against diffusion policies on Push-T, RoboMimic Can, and PointMaze under a shared-backbone comparison protocol that isolates policy-head effects under matched inputs, training budgets, and evaluation settings. Across tasks, liquid policies use roughly half the parameters (4.3M vs. 8.6M), achieve 2.4x lower offline prediction error, and run 1.8 faster at inference. In sample-efficiency experiments spanning 1% to 46.42% of training data, liquid models remain consistently more robust, with especially large gains in low-data and medium-data regimes. Closed-loop results on Push-T and PointMaze are directionally consistent with offline rankings but noisier, indicating that strong offline density modeling helps deployment while not fully determining closed-loop success. Overall, liquid recurrent multimodal policies provide a compact and practical alternative to iterative denoising for imitation learning.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
Parameter Estimation in Stochastic Differential Equations via Wiener Chaos Expansion and Stochastic Gradient Descent
arXiv:2603.27019v1 Announce Type: new
Abstract: This study addresses the inverse problem of parameter estimation for Stochastic Differential Equations (SDEs) by minimizing a regularized discrepancy functional via Stochastic Gradient Descent (SGD). To achieve computational efficiency, we leverage the Wiener Chaos Expansion (WCE), a spectral decomposition technique that projects the stochastic solution onto an orthogonal basis of Hermite polynomials. This transformation effectively maps the stochastic dynamics into a hierarchical system of deterministic functions, termed the \textit{propagator}. By reducing the stochastic inference task to a deterministic optimization problem, our framework circumvents the heavy computational burden and sampling requirements of traditional simulation-based methods like MCMC or MLE. The robustness and scalability of the proposed approach are demonstrated through numerical experiments on various non-linear SDEs, including models for individual biological growth. Results show that the WCE-SGD framework provides accurate parameter recovery even from discrete, noisy observations, offering a significant paradigm shift in the efficient modeling of complex stochastic systems.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
Meta-Harness: End-to-End Optimization of Model Harnesses
arXiv:2603.28052v1 Announce Type: new
Abstract: The performance of large language model (LLM) systems depends not only on model weights, but also on their harness: the code that determines what information to store, retrieve, and present to the model. Yet harnesses are still designed largely by hand, and existing text optimizers are poorly matched to this setting because they compress feedback too aggressively. We introduce Meta-Harness, an outer-loop system that searches over harness code for LLM applications. It uses an agentic proposer that accesses the source code, scores, and execution traces of all prior candidates through a filesystem. On online text classification, Meta-Harness improves over a state-of-the-art context management system by 7.7 points while using 4x fewer context tokens. On retrieval-augmented math reasoning, a single discovered harness improves accuracy on 200 IMO-level problems by 4.7 points on average across five held-out models. On agentic coding, discovered harnesses surpass the best hand-engineered baselines on TerminalBench-2. Together, these results show that richer access to prior experience can enable automated harness engineering.
Fonte: arXiv cs.AI
Theory/Optimization • Score 85
High dimensional theory of two-phase optimizers
arXiv:2603.26954v1 Announce Type: new
Abstract: The trend towards larger training setups has brought a renewed interest in partially asynchronous two-phase optimizers which optimize locally and then synchronize across workers. Additionally, recent work suggests that the one-worker version of one of these algorithms, DiLoCo, shows promising results as a (synchronous) optimizer. Motivated by these studies we present an analysis of LA-DiLoCo, a simple member of the DiLoCo family, on a high-dimensional linear regression problem. We show that the one-worker variant, LA, provides a different tradeoff between signal and noise than SGD, which is beneficial in many scenarios. We also show that the multi-worker version generates more noise than the single worker version, but that this additional noise generation can be ameliorated by appropriate choice of hyperparameters. We conclude with an analysis of SLA -- LA with momentum -- and show that stacking two momentum operators gives an opportunity for acceleration via a non-linear transformation of the "effective'' Hessian spectrum, which is maximized for Nesterov momentum. Altogether our results show that two-phase optimizers represent a fruitful new paradigm for understanding and improving training algorithms.
Fonte: arXiv cs.LG
Theory/Optimization • Score 85
MuonEq: Balancing Before Orthogonalization with Lightweight Equilibration
arXiv:2603.28254v1 Announce Type: cross
Abstract: Orthogonalized-update optimizers such as Muon improve training of matrix-valued parameters, but existing extensions mostly act either after orthogonalization by rescaling updates or before it with heavier whitening-based preconditioners. We introduce {\method}, a lightweight family of pre-orthogonalization equilibration schemes for Muon in three forms: two-sided row/column normalization (RC), row normalization (R), and column normalization (C). These variants rebalance the momentum matrix before finite-step Newton--Schulz using row/column squared-norm statistics and only $\mathcal{O}(m+n)$ auxiliary state. We show that finite-step orthogonalization is governed by input spectral properties, especially stable rank and condition number, and that row/column normalization is a zeroth-order whitening surrogate that removes marginal scale mismatch. For the hidden matrix weights targeted by {\method}, the row-normalized variant R is the natural default and preserves the $\widetilde{\mathcal{O}}(T^{-1/4})$ stationarity guarantee of Muon-type methods. In LLaMA2 pretraining on C4, the default R variant consistently outperforms Muon on 130M and 350M models, yielding faster convergence and lower validation perplexity.
Fonte: arXiv stat.ML
NLP/LLMs • Score 85
KAT-Coder-V2 Technical Report
arXiv:2603.27703v1 Announce Type: new
Abstract: We present KAT-Coder-V2, an agentic coding model developed by the KwaiKAT team at Kuaishou. KAT-Coder-V2 adopts a "Specialize-then-Unify" paradigm that decomposes agentic coding into five expert domains - SWE, WebCoding, Terminal, WebSearch, and General - each undergoing independent supervised fine-tuning and reinforcement learning, before being consolidated into a single model via on-policy distillation. We develop KwaiEnv, a modular infrastructure sustaining tens of thousands of concurrent sandbox instances, and scale RL training along task complexity, intent alignment, and scaffold generalization. We further propose MCLA for stabilizing MoE RL training and Tree Training for eliminating redundant computation over tree-structured trajectories with up to 6.2x speedup. KAT-Coder-V2 achieves 79.6% on SWE-bench Verified (vs. Claude Opus 4.6 at 80.8%), 88.7 on PinchBench (surpassing GLM-5 and MiniMax M2.7), ranks first across all three frontend aesthetics scenarios, and maintains strong generalist scores on Terminal-Bench Hard (46.8) and tau^2-Bench (93.9). Our model is publicly available at https://streamlake.com/product/kat-coder.
Fonte: arXiv cs.CL
MLOps/Systems • Score 85
EpochX: Building the Infrastructure for an Emergent Agent Civilization
arXiv:2603.27304v1 Announce Type: new
Abstract: General-purpose technologies reshape economies less by improving individual tools than by enabling new ways to organize production and coordination. We believe AI agents are approaching a similar inflection point: as foundation models make broad task execution and tool use increasingly accessible, the binding constraint shifts from raw capability to how work is delegated, verified, and rewarded at scale. We introduce EpochX, a credits-native marketplace infrastructure for human-agent production networks. EpochX treats humans and agents as peer participants who can post tasks or claim them. Claimed tasks can be decomposed into subtasks and executed through an explicit delivery workflow with verification and acceptance. Crucially, EpochX is designed so that each completed transaction can produce reusable ecosystem assets, including skills, workflows, execution traces, and distilled experience. These assets are stored with explicit dependency structure, enabling retrieval, composition, and cumulative improvement over time. EpochX also introduces a native credit mechanism to make participation economically viable under real compute costs. Credits lock task bounties, budget delegation, settle rewards upon acceptance, and compensate creators when verified assets are reused. By formalizing the end-to-end transaction model together with its asset and incentive layers, EpochX reframes agentic AI as an organizational design problem: building infrastructures where verifiable work leaves persistent, reusable artifacts, and where value flows support durable human-agent collaboration.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
SCOPE: Tree-based Self-Correcting Online Log Parsing via Syntactic-Semantic Collaboration
arXiv:2603.27247v1 Announce Type: new
Abstract: Log parsing is a critical step for automated log analysis in complex systems. Traditional heuristic-based methods offer high efficiency but are limited in accuracy due to overlooking semantic context. In contrast, recent LLM-based parsers improve accuracy via se mantic understanding but incur high latency from frequent model calls. To address this, we propose SCOPE, the first self-correcting online log parsing method that integrates the strengths of both heuristic and LLM-based paradigms. SCOPE introduces a novel bi-directional tree structure that enables efficient template match ing from both forward and reverse directions, resulting in a higher overall matching rate. Additionally, it adopts a two-stage syntactic semantic collaboration framework: a lightweight NLP model first utilizes part-of-speech (POS) information for syntax-based match ing, while the LLM is selectively invoked as a fallback to handle semantically complex cases when uncertainty remains. This design significantly reduces LLM API usage while maintaining high ac curacy, achieving a balance between efficiency and effectiveness. Extensive evaluations on diverse benchmark datasets show that SCOPE outperforms state-of-the-art methods in both accuracy and efficiency. The implementation and datasets are publicly released to facilitate further research.
Fonte: arXiv cs.CL
Vision • Score 85
Embedding Provenance in Computer Vision Datasets with JSON-LD
arXiv:2603.27348v1 Announce Type: new
Abstract: With the ubiquity of computer vision in industry, the importance of image provenance is becoming more apparent. Provenance provides information about the origin and derivation of some resource, e.g., an image dataset, enabling users to trace data changes to better understand the expected behaviors of downstream models trained on such data. Provenance may also help with data maintenance by ensuring compliance, supporting audits and improving reusability. Typically, if provided, provenance is stored separately, e.g., within a text file, leading to a loss of descriptive information for key details like image capture settings, data preprocessing steps, and model architecture or iteration. Images often lack the information detailing the parameters of their creation or compilation. This paper proposes a novel schema designed to structure image provenance in a manageable and coherent format. The approach utilizes JavaScript Object Notation for Linked Data (JSON-LD), embedding this provenance directly within the image file. This offers two significant benefits: (1) it aligns image descriptions with a robust schema inspired by and linked to established standards, and (2) it ensures that provenance remains intrinsically tied to images, preventing loss of information and enhancing system qualities, e.g., maintainability and adaptability. This approach emphasizes maintaining the direct connection between vision resources and their provenance.
Fonte: arXiv cs.LG
MLOps/Systems • Score 85
Throughput Optimization as a Strategic Lever in Large-Scale AI Systems: Evidence from Dataloader and Memory Profiling Innovations
arXiv:2603.26823v1 Announce Type: new
Abstract: The development of large-scale foundation models, particularly Large Language Models (LLMs), is constrained by significant computational and memory bottlenecks. These challenges elevate throughput optimization from a mere engineering task to a critical strategic lever, directly influencing training time, operational cost, and the feasible scale of next-generation models. This paper synthesizes evidence from recent academic and industry innovations to analyze key advancements in training efficiency. We examine architectural solutions to dataloader bottlenecks, such as the OVERLORD framework, which has demonstrated a 4.5% improvement in end-to-end training throughput. We investigate memory optimization techniques designed to overcome the GPU memory wall, including CPU offloading strategies like DeepSpeed's ZeRO-Offload, which enable the training of models far exceeding single-accelerator capacity. Furthermore, we explore the growing importance of compiler-centric optimizations, exemplified by Triton-distributed, which enables the joint optimization of computation, memory, and communication for substantial performance gains. The analysis is contextualized by advanced profiling tools and hardware characterization studies that identify and mitigate previously overlooked overheads like Dynamic Voltage and Frequency Scaling (DVFS). Findings indicate that a holistic, system-level approach, integrating innovations across data pipelines, memory management, network fabrics, and compiler technologies, is essential for accelerating AI development, managing costs, and pushing the boundaries of model scale.
Fonte: arXiv cs.LG
RL • Score 85
RG-TTA: Regime-Guided Meta-Control for Test-Time Adaptation in Streaming Time Series
arXiv:2603.27814v1 Announce Type: cross
Abstract: Test-time adaptation (TTA) enables neural forecasters to adapt to distribution shifts in streaming time series, but existing methods apply the same adaptation intensity regardless of the nature of the shift. We propose Regime-Guided Test-Time Adaptation (RG-TTA), a meta-controller that continuously modulates adaptation intensity based on distributional similarity to previously-seen regimes. Using an ensemble of Kolmogorov-Smirnov, Wasserstein-1, feature-distance, and variance-ratio metrics, RG-TTA computes a similarity score for each incoming batch and uses it to (i) smoothly scale the learning rate -- more aggressive for novel distributions, conservative for familiar ones -- and (ii) control gradient effort via loss-driven early stopping rather than fixed budgets, allowing the system to allocate exactly the effort each batch requires. As a supplementary mechanism, RG-TTA gates checkpoint reuse from a regime memory, loading stored specialist models only when they demonstrably outperform the current model (loss improvement >= 30%). RG-TTA is model-agnostic and strategy-composable: it wraps any forecaster exposing train/predict/save/load interfaces and enhances any gradient-based TTA method. We demonstrate three compositions -- RG-TTA, RG-EWC, and RG-DynaTTA -- and evaluate 6 update policies (3 baselines + 3 regime-guided variants) across 4 compact architectures (GRU, iTransformer, PatchTST, DLinear), 14 datasets (6 real-world multivariate benchmarks + 8 synthetic regime scenarios), and 4 forecast horizons (96, 192, 336, 720) under a streaming evaluation protocol with 3 random seeds (672 experiments total). Regime-guided policies achieve the lowest MSE in 156 of 224 seed-averaged experiments (69.6%), with RG-EWC winning 30.4% and RG-TTA winning 29.0%. Overall, RG-TTA reduces MSE by 5.7% vs TTA while running 5.5% faster; RG-EWC reduces MSE by 14.1% vs standalone EWC.
Fonte: arXiv stat.ML
MLOps/Systems • Score 85
Calorimeter Shower Superresolution with Conditional Normalizing Flows: Implementation and Statistical Evaluation
arXiv:2603.26813v1 Announce Type: cross
Abstract: In High Energy Physics, detailed calorimeter simulations and reconstructions are essential for accurate energy measurements and particle identification, but their high granularity makes them computationally expensive. Developing data-driven techniques capable of recovering fine-grained information from coarser readouts, a task known as calorimeter superresolution, offers a promising way to reduce both computational and hardware costs while preserving detector performance. This thesis investigates whether a generative model originally designed for fast simulation can be effectively applied to calorimeter superresolution. Specifically, the model proposed in arXiv:2308.11700 is re-implemented independently and trained on the CaloChallenge 2022 dataset based on the Geant4 Par04 calorimeter geometry. Finally, the model's performance is assessed through a rigorous statistical evaluation framework, following the methodology introduced in arXiv:2409.16336, to quantitatively test its ability to reproduce the reference distributions.
Fonte: arXiv stat.ML
NLP/LLMs • Score 88
Robust Batch-Level Query Routing for Large Language Models under Cost and Capacity Constraints
arXiv:2603.26796v1 Announce Type: new
Abstract: We study the problem of routing queries to large language models (LLMs) under cost, GPU resources, and concurrency constraints. Prior per-query routing methods often fail to control batch-level cost, especially under non-uniform or adversarial batching. To address this, we propose a batch-level, resource-aware routing framework that jointly optimizes model assignment for each batch while respecting cost and model capacity limits. We further introduce a robust variant that accounts for uncertainty in predicted LLM performance, along with an offline instance allocation procedure that balances quality and throughput across multiple models. Experiments on two multi-task LLM benchmarks show that robustness improves accuracy by 1-14% over non-robust counterparts (depending on the performance estimator), batch-level routing outperforms per-query methods by up to 24% under adversarial batching, and optimized instance allocation yields additional gains of up to 3% compared to a non-optimized allocation, all while strictly controlling cost and GPU resource constraints.
Fonte: arXiv cs.LG
RL • Score 85
Bitboard version of Tetris AI
arXiv:2603.26765v1 Announce Type: new
Abstract: The efficiency of game engines and policy optimization algorithms is crucial for training reinforcement learning (RL) agents in complex sequential decision-making tasks, such as Tetris. Existing Tetris implementations suffer from low simulation speeds, suboptimal state evaluation, and inefficient training paradigms, limiting their utility for large-scale RL research. To address these limitations, this paper proposes a high-performance Tetris AI framework based on bitboard optimization and improved RL algorithms. First, we redesign the Tetris game board and tetrominoes using bitboard representations, leveraging bitwise operations to accelerate core processes (e.g., collision detection, line clearing, and Dellacherie-Thiery Features extraction) and achieve a 53-fold speedup compared to OpenAI Gym-Tetris. Second, we introduce an afterstate-evaluating actor network that simplifies state value estimation by leveraging Tetris afterstate property, outperforming traditional action-value networks with fewer parameters. Third, we propose a buffer-optimized Proximal Policy Optimization (PPO) algorithm that balances sampling and update efficiency, achieving an average score of 3,829 on 10x10 grids within 3 minutes. Additionally, we develop a Python-Java interface compliant with the OpenAI Gym standard, enabling seamless integration with modern RL frameworks. Experimental results demonstrate that our framework enhances Tetris's utility as an RL benchmark by bridging low-level bitboard optimizations with high-level AI strategies, providing a sample-efficient and computationally lightweight solution for scalable sequential decision-making research.
Fonte: arXiv cs.AI
NLP/LLMs • Score 85
Self-evolving AI agents for protein discovery and directed evolution
arXiv:2603.27303v1 Announce Type: new
Abstract: Protein scientific discovery is bottlenecked by the manual orchestration of information and algorithms, while general agents are insufficient in complex domain projects. VenusFactory2 provides an autonomous framework that shifts from static tool usage to dynamic workflow synthesis via a self-evolving multi-agent infrastructure to address protein-related demands. It outperforms a set of well-known agents on the VenusAgentEval benchmark, and autonomously organizes the discovery and optimization of proteins from a single natural language prompt.
Fonte: arXiv cs.AI