Search: “arxiv-cs-lg”

Variance-Optimal State Resampling for Reinforcement Learning with Verifiable Rewards

arXiv:2610.15888v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards spends much of its rollout budget on states that yield no learning signal. We recast this as a question a

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

VH-GPO: Verified Handoff for LLM Agent Reinforcement Learning

DGX agent

arXiv:2610.14000v1 Announce Type: new Abstract: Group-relative reinforcement learning methods such as GRPO train large language model (LLM) agents by comparing rollouts for the same task. When every t

safetyarxiv-cs-lg

14 Oct 2026

Vision-Predictor-Action: Towards Efficient Embodied AI via Neuro-Symbolic JEPA Predictors

arXiv:2610.15556v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) models commonly generate robot actions token by token through Large Language Model (LLM) decoders, causing inference late

Visualizing the World Model in an Agent's Value Function

arXiv:2610.14186v1 Announce Type: cross Abstract: Could model-free reinforcement learning algorithms actually be learning a model of the environment? In this work, we apply counterfactual intervention

VitaBench: Memory Lifecycles Beyond Retrieval

arXiv:2610.14731v1 Announce Type: new Abstract: Long-term memory benchmarks for language agents primarily evaluate whether information from earlier interactions can be retrieved and used when relevant

Applications

Watermarking Continuous Tabular Data Generation

arXiv:2405.14018v2 Announce Type: replace-cross Abstract: In this paper, we introduce a simple yet effective tabular data watermarking mechanism with statistical guarantees. We show theoretically that

applicationsarxiv-cs-lg

Tutorials

Weighting-Based Strategies for Classification: Analysis and Local Peak Dynamics with CART Trees

arXiv:2610.18037v1 Announce Type: cross Abstract: Tabular learning tasks often suffer from class imbalance, which is often addressed via weighting-based strategies. In this work, we compare the naive

tutorialsarxiv-cs-lg

What a Data Release Must Preserve for Off-Policy Evaluation

arXiv:2610.15283v1 Announce Type: cross Abstract: In off-policy evaluation, merging decision-time histories that matter differently for a specified policy value or comparison can make that evaluation

What Can Language Models Inherit from Solomonoff Induction? Limits, Attainable Properties, and What Served Conditionals Reveal

arXiv:2610.14007v1 Announce Type: new Abstract: A familiar line of reasoning runs as follows: language models are flexible approximators trained under a strictly proper loss, so scale brings them clos

Tutorials

What DNA Foundation Models Learn Beyond Sequence Composition

arXiv:2610.15711v1 Announce Type: new Abstract: Do genomic foundation models (FMs) provide predictive information beyond simple sequence composition? We evaluate frozen representations from five DNA F

tutorialsarxiv-cs-lg

What Do Sparse Autoencoder Feature Discovery Scores Identify? From Activations to Interventions

arXiv:2610.15850v1 Announce Type: new Abstract: Sparse autoencoder (SAE) features are typically discovered from how their activations co-vary with behaviour, yet they are often used through interventi

What Drives AI Citations? Content, Retrieval and Model Knowledge in Generative Engine Optimization (GEO)

arXiv:2610.15033v1 Announce Type: cross Abstract: Generative search engines such as ChatGPT, Gemini and Perplexity now decide which Web pages users see as sources, and a growing industry of generative

What is the Value of Censored Data? An Exact Analysis for the Data-driven Newsvendor

arXiv:2602.16842v2 Announce Type: replace Abstract: We study the offline data-driven newsvendor problem when historical demand is censored by past inventory levels. We introduce a class of policies fo

What Matters in Quality-Aware VLA Pretraining: A Benchmark and Empirical Study

arXiv:2610.18869v1 Announce Type: cross Abstract: How should demonstration quality be used during vision-language-action (VLA) pretraining to improve performance after target-task fine-tuning? We comp

What Survives When Activation Steering Is Used as Memory

arXiv:2610.14958v1 Announce Type: cross Abstract: Activation steering has recently been used as a memory for skills: past experiences of an agent are stored as steering vectors and retrieved later. Fa

When do explicit penalties improve flow matching?

arXiv:2610.15689v1 Announce Type: cross Abstract: Flow matching can learn structure directly from data, raising the question of when additional constraints improve generation. We study penalties that

When Do Learned Blocks Preserve Causal Meaning? Block Identifiability and Constructive Causal Abstractions

arXiv:2610.18155v1 Announce Type: new Abstract: Standard identifiability theory assumes one-dimensional latent factors, but learned representations can encode latent information across multidimensiona

When Does Compensating Inherited Quantization Error Help? Mean-Square Guarantees for Layer-Wise Quantization and Their Limits

arXiv:2610.16020v1 Announce Type: new Abstract: Layer-wise post-training quantization methods let each layer compensate for the quantization error inherited from earlier layers. We ask under what assu

When Does Opponent Modeling Pay Off? Spatial Inference Stabilises Reinforcement Learning under Partial Observability

arXiv:2610.18229v1 Announce Type: cross Abstract: Deep reinforcement learning agents in partially observable competitive games must act on beliefs about an unseen opponent's intent. We study a spatial

When Does Trimming Help Conformal Prediction? A Retained-Law Diagnostic under Calibration Contamination

arXiv:2605.06204v2 Announce Type: replace-cross Abstract: Trimming suspicious calibration points is a common response to contamination in conformal prediction. Its effect on clean-target coverage, how

When Dynamic Scaling Fails: A Scale-State Feedback Governs Recurrent-State Quantization in Linear-Attention and SSM LLMs

arXiv:2610.15139v1 Announce Type: new Abstract: Linear-attention and state-space LLMs carry long context in a fixed-size recurrent state, and quantizing that state is the natural way to cut its cost d

When LLM Agents See a Conflict but Still Proceed: Cross-Domain Dependencies in Network Slicing

arXiv:2610.17118v1 Announce Type: cross Abstract: Future AI-native networks are being designed to apply autonomous agents built on large language models (LLMs) in network management through intents. W

Hardware

When Log Parsing Removes What Matters: Detecting Failures in Distributed GPU Workloads

arXiv:2610.14009v1 Announce Type: new Abstract: Large AI models increasingly rely on distributed GPU execution, yet it remains unclear whether established failure analysis methods transfer effectively

hardwarearxiv-cs-lg

Local Ai

When Missingness Follows the Graph: Beyond Feature Propagation for Missing Features

arXiv:2610.18113v1 Announce Type: new Abstract: Graph Neural Networks (GNNs) commonly assume that node features are fully observed, despite missing features being pervasive in real-world applications.

local-aiarxiv-cs-lg

When More Memory Fails: Hidden Feasibility Holes in Dynamic Tensor Rematerialization

arXiv:2610.14369v1 Announce Type: new Abstract: Dynamic Tensor Rematerialization (DTR) trains neural networks under a memory budget by evicting tensors and recomputing them on demand. A natural expect

Applications

When More Objectives Make Identification Easier: The Role of Preferences and Feedback

arXiv:2610.18577v1 Announce Type: new Abstract: Best-arm identification (BAI) aims to identify the optimal arm using as few samples as possible. While most existing work focuses on a single objective,

applicationsarxiv-cs-lg

When Participant Disjoint Validation Is Not Enough to Establish Generalization in Parkinson's Gait Classification

arXiv:2610.16220v1 Announce Type: new Abstract: Participant-disjoint cross-validation prevents direct participant overlap, but it may still produce optimistic estimates when acquisition site and diagn

When Repeated Measurements Hurt Tabular Foundation Models

arXiv:2610.16801v1 Announce Type: cross Abstract: Tabular foundation models (TFMs) achieve state-of-the-art performance on IID benchmarks, but how do they respond to repeated measurements at the same

When Sparsity Breaks Connectivity: Functional Reachability and Guided Topology Dynamics in Extreme Sparse Training

arXiv:2610.17425v1 Announce Type: new Abstract: Dynamic Sparse Training (DST) periodically prunes and regrows connections, inducing a discrete dynamical mechanism over network topologies. At extreme s

Where Small Mistakes Matter: Quantifying State Criticality for Imitation Learning

arXiv:2610.17709v1 Announce Type: new Abstract: Imitation learning typically treats every demonstration state as equally important, optimizing a behavior-cloning objective uniformly across the dataset

Which Geometry on Which Layer? Measured Sign-Spectral Assignment in Mixed-Optimizer Training

arXiv:2610.17285v1 Announce Type: new Abstract: Modern neural-network training uses mixed recipes, assigning AdamW-like or sign-based updates to embedding and output layers and spectral updates to hid

Which Optimizer Leaks Most? Optimizer Implicit Bias as a Training-Data Privacy Controller

arXiv:2610.18294v1 Announce Type: new Abstract: Releasing a neural network checkpoint can leak information about its training data, and existing reconstruction results rely on gradient-descent trainin

Who Improves the Robot? Toward Physical Recursive Self-Improvement

arXiv:2610.18842v1 Announce Type: cross Abstract: Robots are increasingly autonomous in task execution, yet their improvement remains largely human-driven. This study investigates Physical Recursive S

Who Sets the Pace of Judge Hacking? Judge Capability versus the Policy-Judge Gap

arXiv:2610.15778v1 Announce Type: new Abstract: When a policy is trained with GRPO against a frozen, reference-free LLM judge, the judge accepts more of its answers, but in a post-hoc analysis we dete

WildTrace: Evaluating Training Data Detection in Realistic Large-Scale LLM Post-Training

arXiv:2610.16152v1 Announce Type: new Abstract: Although post-training significantly enhances model performance, the use of massive undisclosed data raises concerns about benchmark contamination and c

WisdomNet: trustable machine learning toward error-free classification

arXiv:2610.15299v1 Announce Type: new Abstract: Misclassification is a critical problem in many machine learning applications. Since even the classifier models with high accuracy (e.g., > 95%) still i

WorkRover: Scaling Long-Horizon Cowork Agent Training Via Recursive Environment Evolution

arXiv:2610.18067v1 Announce Type: cross Abstract: Scaling interactive environments is important for training long-horizon agents through reinforcement learning. These environments often require furthe

World Modeling through Spectral Alignment

arXiv:2610.15429v1 Announce Type: new Abstract: How do we tell a world model what information to preserve? We study this question through the lens of spectral representation learning. We introduce Spe

World2TS: Real-World Event Context for Time-Series Forecasting

arXiv:2610.14861v1 Announce Type: new Abstract: Time-series forecasting plays a vital role in a wide range of real-world applications. Recent approaches have achieved strong performance by modeling te

WUSPruner: Second-Moment-Optimal Transforms for Structured Pruning

arXiv:2610.15394v1 Announce Type: new Abstract: Structured pruning is typically formulated as identifying which model dimensions to remove, often within the model's original representation. However, t

XForecast: The KDD 2026 Finance Day Challenge on News-Grounded Financial Forecasting

arXiv:2610.16131v1 Announce Type: new Abstract: Financial forecasting has traditionally relied on numerical time-series (TS) signals, while textual information such as news remains largely underutiliz

Zeroth-Order Fine-Tuning of LLMs in Learnable Low-Rank Subspaces

arXiv:2610.15225v1 Announce Type: new Abstract: Zeroth-order (ZO) optimization enables memory-efficient fine-tuning of large language models (LLMs). However, ZO optimization suffers from high variance

ZOIF: Zeroth-Order Influence Function Estimation for LLM Data Selection Using Only Forward Passes

arXiv:2610.15336v1 Announce Type: new Abstract: Influence functions enable principled data selection for large language models (e.g., identifying influential training samples for model editing, debugg

zPocket: Better Pocket Finding with Frozen Co-Folding Features

arXiv:2610.17696v1 Announce Type: new Abstract: Identifying protein-ligand binding sites (pockets) is important for understanding protein function and developing drugs. However, no public method has l

A Categorical Atlas of Parameter-Efficient Fine-Tuning: Cones, Orbits, and Gauges

arXiv:2610.12728v1 Announce Type: new Abstract: Parameter-efficient fine-tuning has hundreds of methods, usually compared by parameter count and benchmark score. We compare them as maps. A pretrained

A Critical Look at Power Sampling when Test Time Sharpening LLMs

arXiv:2610.13653v1 Announce Type: cross Abstract: Reinforcement learning with verifiable rewards (RLVR) has substantially improved LLM reasoning, yet growing evidence suggests it mainly performs distr

A Distributional Framework for Behavioral Characterization and Risk Management of Large Language Models

arXiv:2610.12718v1 Announce Type: cross Abstract: Large language models are stochastic by nature. Repeated executions under the same experimental conditions may produce different responses. We propose

A Generalizable Machine Learning Framework for Half-Heusler Thermoelectrics

arXiv:2602.01149v2 Announce Type: replace-cross Abstract: Machine learning (ML) has emerged as a powerful approach for accelerating the discovery of thermoelectric (TE) materials, yet many models suff