AI news archive.
10,659 articles filed under Research, newest first, from the outlets listed on the sources page.
- arXiv cs.AIresearchTrust Mechanisms for Offline Generative AI at the Edge: A Connectivity-Regime Framework and an Empirical Study on a Raspberry Pi 5
arXiv:2610.33732v1 Announce Type: cross Abstract: Generative AI and large language models (LLMs) are moving into the Internet of Things, yet most deployments assume cloud connectivity, and many IoT settings have intermit…
- arXiv cs.AIresearchA Distribution-Matching Perspective on LLM Training Algorithms
arXiv:2610.33740v1 Announce Type: cross Abstract: In this work, we propose a unified framework for training algorithms across different stages of modern large language model development, based on the perspective of distr…
- arXiv cs.AIresearchPOD-RoPE: Rotary Position Encoding over Physical Observation Distributions for Embodied AI
arXiv:2610.33745v2 Announce Type: replace-cross Abstract: Embodied perception requires positional representations that capture both physical geometry and the spatial and temporal extent of observations. We introduce POD-…
- arXiv cs.AIresearchCFL-aware edge design for graph neural solvers of hyperbolic transport
arXiv:2610.33769v1 Announce Type: cross Abstract: Hyperbolic partial differential equations (PDEs) transport information at finite speed along characteristics, often producing shocks and discontinuities that are challeng…
- arXiv cs.AIresearchTriageRAG: Failure-Aware Self-Training for Adapting Retrieval-Augmented Language Models to Specialized Domains
arXiv:2610.33776v1 Announce Type: cross Abstract: Retrieval-augmented generation (RAG) lets large language models (LLMs) answer knowledge-intensive questions from retrieved evidence, but adapting a general-purpose RAG sy…
- arXiv cs.AIresearchRoboReason: Boosting Robotics-Relevant VLM Reasoning Through Scalable Data Generation
arXiv:2610.33778v1 Announce Type: cross Abstract: Vision-Language Models (VLMs) have demonstrated impressive performance when integrated into robot manipulation pipelines, but real-world evaluation of VLM capabilities on…
- arXiv cs.AIresearchOne Scan is Enough: Demonstration-Free Adaptation for Language-Guided Navigation
arXiv:2610.33794v1 Announce Type: cross Abstract: Deploying a language-guided navigation policy in a new building usually means watching it fail. Public-benchmark VLAs degrade scene by scene as layout, appearance, and la…
- arXiv cs.AIresearchProMo3D: Probing Motion Cues in Frozen 3D Foundation Models
arXiv:2610.33802v1 Announce Type: cross Abstract: Feed-forward 3D foundation models have recently reshaped multi-view reconstruction, yet dynamic 4D understanding still often relies on large-scale motion supervision or h…
- arXiv cs.AIresearchLearning What to Share: Expert Specialization for Time-Series Forecasting
arXiv:2610.33822v1 Announce Type: cross Abstract: A central challenge in multivariate forecasting is deciding what to share across channels and where to specialize. Mixture-of-experts models offer flexible sharing, yet i…
- arXiv cs.AIresearchAdapting DNNs to Embrace Voltage Scaling Induced Errors via Accumulation-Aware Quantization
arXiv:2610.33824v1 Announce Type: cross Abstract: The pervasive deployment of Deep Neural Networks (DNNs) across the computing spectrum has brought growing attention to energy efficiency, alongside raw performance. While…
- arXiv cs.AIresearchEEGBiFormer: Real-Time EEG Attention Classification via Hybrid CNN-Transformer with EEG-Informed Auditory Protocol Validation
arXiv:2610.33832v1 Announce Type: cross Abstract: Sustained attention is critical for academic performance, yet existing auditory support tools are static and not grounded in real-time neural evidence. This paper present…
- arXiv cs.AIresearchIntervention-Certified Amplification: A Validity Gate for $\sigma$-Weighted Post-Training
arXiv:2610.33882v1 Announce Type: cross Abstract: Hybrid post-training often mixes supervised fine-tuning (SFT) and reinforcement learning (RL), using training-side signals to set each sample's balance. Such signals char…
- arXiv cs.AIresearchCounterfactual Test-Time Augmentation for Shortcut Removal in Language Models
arXiv:2610.33904v1 Announce Type: cross Abstract: Large language models (LLMs) can exploit shortcuts, i.e., correlations present in training data that may not hold at test time. This can lead to poor predictions when cor…
- arXiv cs.AIresearchRank-1 Weight-Corrected Trainable Weight Averaging for Out-of-Distribution Detection
arXiv:2610.33909v1 Announce Type: cross Abstract: Out-of-distribution (OOD) detection identifies inputs outside the training distribution. Rank-based analyses such as RankFeat and RankWeight suggest that OOD overconfiden…
- arXiv cs.AIresearchMultitask Interactions in Recurrent Dynamics across 100 Cognitive Tasks
arXiv:2610.33926v1 Announce Type: cross Abstract: The brain continually learns tasks with very different computational demands. Recurrent neural networks (RNNs) have been used as dynamical surrogates for studying task so…
- arXiv cs.AIresearchWhat Makes a Question Curious? Measuring Model Curiosity in Human-LLM Interactions
arXiv:2610.33950v1 Announce Type: cross Abstract: Large language models increasingly act as partners in everyday communication. In these interactions, useful assistance may require not only answering, but also asking the…
- arXiv cs.AIresearchBeyond Belief: Evaluating and Mitigating LLM Susceptibility to Misinformation
arXiv:2610.33962v1 Announce Type: cross Abstract: Tool-using large language models (LLMs) must decide whether external information warrants action. We investigate whether their expressed credibility judgments reliably in…
- arXiv cs.AIresearchQECGBERT: A Scalable ECG Foundation Model Based on Self-Supervised Language Learning via VQ-VAE Waveform Tokenization
arXiv:2610.33964v1 Announce Type: cross Abstract: AI foundation models play a pivotal role in advancing heart disease diagnosis from electrocardiogram (ECG) data, particularly for rare diseases where the lack of sufficie…
- arXiv cs.AIresearchHierarchical Cooperative GFlowNets
arXiv:2610.33968v2 Announce Type: replace-cross Abstract: Generative Flow Networks (GFlowNets) are trained to sample objects from compositional spaces with probability proportional to a reward function, but ineffective e…
- arXiv cs.AIresearchSteering the Teacher: Improving On-Policy Self-Distillation for Thinking Models
arXiv:2610.34005v1 Announce Type: cross Abstract: On-policy self-distillation (OPSD) trains a student on its own reasoning trajectories using token-level supervision from a reference-conditioned teacher initialized from…
- arXiv cs.AIresearchBriLLM-MetaPred: A Representation-Prediction Unity Language Model with Adaptive Self-Measurement
arXiv:2610.34018v1 Announce Type: cross Abstract: We present BriLLM-MetaPred, the first engineering implementation of the MetaPred machine-learning framework established by a mathematical theory of AGI, instantiated for…
- arXiv cs.AIresearchDeep Scientific Reasoning under Physical Constraints: Structure-Aware Spectrum Prediction
arXiv:2610.34021v1 Announce Type: cross Abstract: Structured scientific spectra encode rich physical information while obeying hard constraints, such as conservation laws and characteristic spectral geometry. Accurate pr…
- arXiv cs.AIresearchA Sign Law for Entropy Drift under Sampler Mismatch in RLVR
arXiv:2610.34044v2 Announce Type: replace-cross Abstract: In reinforcement learning with verifiable rewards (RLVR), whether entropy rises or falls can depend on choices that matter little under exact sampling, such as wh…
- arXiv cs.AIresearchExtended LASMP: Language Aided Subset Sampling Based Motion Planner for Drone Navigation
arXiv:2610.34054v1 Announce Type: cross Abstract: This paper proposes a command-aligned sampling distribution that incorporates navigation commands directly into the exploration process to improve the efficiency of sampl…
- arXiv cs.AIresearchIdentically Trained Language Models Break on Different Jailbreaks
arXiv:2610.34074v1 Announce Type: cross Abstract: Jailbreak benchmarks test one checkpoint and treat which jailbreaks break it as a property of its pretraining data and safety training. We ask whether another pretraining…
- arXiv cs.AIresearchT2D: Target-to-Draft Model Pruning & Adaptation for Efficient Speculative Decoding
arXiv:2610.34103v1 Announce Type: cross Abstract: Speculative decoding significantly accelerates autoregressive text generation in Large Language Models (LLMs), but its efficacy depends largely on two key factors: the di…
- arXiv cs.AIresearchHumanoidNDM: Closing the Training-Deployment Gap of Motion Tracking via Humanoid Neural Dynamics Model
arXiv:2610.34126v1 Announce Type: cross Abstract: Humanoid robots have recently achieved remarkable success in motion tracking. However, during real-world deployment, robots inevitably encounter dynamics gaps arising fro…
- arXiv cs.AIresearchDecoupled Prune-and-Reconstruct: Training on a Compact Core with Double-Layer Sparsification
arXiv:2610.34140v1 Announce Type: cross Abstract: On million-node graphs, the memory wall arrives before the compute wall: a GNN that never fits in VRAM cannot deliver the all-node inference its accuracy promises. The st…
- arXiv cs.AIresearchRethinking Retrieval in the Long-Context Era: Training-Free Semantic Trees for Adaptive Context Construction
arXiv:2610.34146v1 Announce Type: cross Abstract: Long-context language models can increasingly process entire documents, making full-context prompting a practical alternative to passage retrieval. However, greater capac…
- arXiv cs.AIresearchSingle-Cycle Battery Remaining Useful Life Prediction with Compressed Continuous Wavelet Features and Deep Neural Networks
arXiv:2610.34159v1 Announce Type: cross Abstract: Lithium-ion batteries are widely used in industrial applications, but their performance gradually degrades over time, making accurate lifetime prediction important for sa…
- arXiv cs.AIresearchWrite What Matters: Disentangling Token Selection from Slot Allocation in Recurrent Memory
arXiv:2610.34180v1 Announce Type: cross Abstract: Long-sequence models must retain earlier information. Full attention keeps all past tokens but grows expensive, and recurrent-memory models compress history into fixed sl…
- arXiv cs.AIresearchEntropic Control for Homeostatic Optimization: Pre-empting Catastrophic Forgetting via Dynamic Feedback
arXiv:2610.34182v1 Announce Type: cross Abstract: Supervised Fine-Tuning (SFT) of LLMs suffers from catastrophic forgetting, and prevailing mitigation strategies struggle to reconcile the plasticity-stability dilemma. To…
- arXiv cs.AIresearchSRA: Selective Residual Adaptation for LLM Reasoning Enhancement
arXiv:2610.34199v1 Announce Type: cross Abstract: Knowledge distillation transfers reasoning capabilities from large language models to lightweight students. However, distilled small models still struggle with multi-step…
- arXiv cs.AIresearchCORMA: Constraint-Routed Decision Mamba for Generalizable Multi-Constraint Offline Safe Reinforcement Learning
arXiv:2610.34208v1 Announce Type: cross Abstract: Offline safe reinforcement learning (RL) learns safety-aware policies from static datasets while satisfying safety constraints during deployment. Motivated by the strong…
- arXiv cs.AIresearchStaX-Net: Stackelberg Interleaved Asymmetric Network for Cross-Modality Medical Image Translation
arXiv:2610.34214v1 Announce Type: cross Abstract: Cross-modality medical image translation is challenged by anatomically inconsistent synthesis that can compromise clinical diagnosis. We propose StaX-Net, a Stackelberg i…
- arXiv cs.AIresearchRecovering Business Rules as Cascade Lookups for Enterprise Field Prediction
arXiv:2610.34258v1 Announce Type: cross Abstract: Enterprise sales-order systems rely on manually maintained configuration rules to determine fields such as payment terms, shipping points, and sales ownership. When these…
- arXiv cs.AIresearchDoubleFree:Exploiting Exploration History for Rollout-Free Multi-Task LLMs Training
arXiv:2610.34272v1 Announce Type: cross Abstract: Multi-task post-training can develop domain specialists through independent reinforcement learning (RL) and consolidate their capabilities through multi-teacher on-policy…
- arXiv cs.AIresearchRepairing Broken Access Control in Web Applications with LLMs: An Empirical Benchmark Study
arXiv:2610.34290v1 Announce Type: cross Abstract: Broken access control (BAC) ranks among the most severe web-application vulnerability classes, yet its automated repair remains largely unexplored. Unlike memory-safety v…
- arXiv cs.AIresearchAutomated, auditable dating of the Babylonian Astronomical Diaries with language models and astronomical tools
arXiv:2610.34297v1 Announce Type: cross Abstract: The Babylonian Astronomical Diaries record the sky over nearly six centuries. Many tablets are broken and have lost their date formula. Experts dated many of them one tab…
- arXiv cs.AIresearchDeveloper Task Delegation to LLMs and the Trust Factor: A Mixed Methods Study
arXiv:2610.34317v1 Announce Type: cross Abstract: With the rise of Large Language Models (LLMs), software developers are presented with a powerful new approach: delegating software development tasks to LLMs. Existing stu…
- arXiv cs.AIresearchSHIFT-JEV: Certified Abstention and Gain-Gated Continual Adaptation for LLM-Based Decision Models under Distribution Shift
arXiv:2610.34329v1 Announce Type: cross Abstract: Large language models (LLMs) can act as decision models that return a probability distribution over a fixed set of options, but deployed decision models face domain, labe…
- arXiv cs.AIresearchCoLink: Global Compression and Knowledge Linking for Multi-Turn Dialogue Modeling
arXiv:2610.34335v1 Announce Type: cross Abstract: Extracting salient information from long-horizon dialogue histories remains a significant challenge for Large Language Models (LLMs) due to the restricted context windows…
- arXiv cs.AIresearchImproving the Chain-of-Thought Compressibility of Pruned Reasoning Models via Long--Short Contrastive Pruning
arXiv:2610.34336v1 Announce Type: cross Abstract: Large reasoning models are costly to serve because they are large and because they generate long chains of thought (CoT). Weight pruning and CoT compression address these…
- arXiv cs.AIresearchXing USP: A Cross-Family Semantic Contract for Information Extraction and Intent-Slot Parsing
arXiv:2610.34343v1 Announce Type: cross Abstract: Structured semantic parsing is typically judged by whether the correct category is assigned, which leaves value recovery and instance-level associations unexamined: a sys…
- arXiv cs.AIresearchComBeautify: Leveraging Code Change Summaries for Conventional Commit Generation
arXiv:2610.34374v1 Announce Type: cross Abstract: Modern software development relies on version control systems such as Git to track changes and facilitate collaboration. While commit messages are crucial for conveying d…
- arXiv cs.AIresearchWhether or Where? Tail Calibration Signals Finite-Budget Reward Hacking in Best-of-N Selection
arXiv:2610.34375v1 Announce Type: cross Abstract: Best-of-N (BoN) inference returns the candidate with the highest proxy reward, so a larger $N$ tests the proxy's extreme tail more than its average ranking quality. We sh…
- arXiv cs.AIresearchDynamic Multi-Step Decoding for Time-Series Foundation Models
arXiv:2610.34409v1 Announce Type: cross Abstract: Time-series forecasting models typically use a fixed decoding strategy at inference time: either one-shot prediction of the full horizon or autoregressive rollout over at…
- arXiv cs.AIresearchIn-Context Adapter Generation for Compositional Generalization in Large Audio Language Models
arXiv:2610.34413v1 Announce Type: cross Abstract: Large audio language models (LALMs) have achieved strong performance on standard speech and audio understanding tasks, yet their ability to systematically generalize to n…
- arXiv cs.AIresearchSpatial Audio Generation and Editing with Executable Plans and Self-Distilling Transfusion
arXiv:2610.34415v1 Announce Type: cross Abstract: Spatial audio authoring coordinates source decisions with soundfield synthesis. For editing, plans specify what to change and retain, while reference audio supplies furth…
- arXiv cs.AIresearchEfficient Proof Generation for Lean Statements with Multiple Candidate Answers
arXiv:2610.34490v1 Announce Type: cross Abstract: In formal systems such as Lean 4, efficiently finding answers to mathematical problems and generating verifiable proofs is an important task in automated mathematical rea…