AI news archive.
10,659 articles filed under Research, newest first, from the outlets listed on the sources page.
- arXiv cs.AIresearchMesh-Native Tokenisation: Retrofitting Foundation Models to Unstructured Data
arXiv:2610.12618v1 Announce Type: cross Abstract: Physics foundation models are an increasingly popular approach for building versatile emulators for spatiotemporal systems. They require tokenisation layers that lift phy…
- arXiv cs.AIresearchMachine Learning and Time-Series Foundation Models for Day-Ahead Electricity Forecasting in a University Building
arXiv:2610.12638v1 Announce Type: cross Abstract: Day-ahead electricity forecasts can help building operators schedule heating, ventilation, and other services around anticipated demand. Choosing a forecasting model also…
- arXiv cs.AIresearchCancerous Backdoor Attack: Gradual Activation via Invisible Tokens and A Stage Controller
arXiv:2610.12644v1 Announce Type: cross Abstract: Backdoor Attacks on Large Language Models (LLMs) typically follow a trigger-response pattern, where satisfying a trigger condition activates a single terminal behavior. H…
- arXiv cs.AIresearchKronecker Product Attention: Memory-Efficient KV Compression via Structured Factorization
arXiv:2610.12657v1 Announce Type: cross Abstract: Efficient autoregressive generation is increasingly limited by the memory footprint of the key-value (KV) cache, which grows linearly with sequence length, batch size, an…
- arXiv cs.AIresearchIncremental Retrieval-Augmented Generation for Human-in-the-Loop Decision Support
arXiv:2610.12666v1 Announce Type: cross Abstract: We introduce iRAG, a human-in-the-loop decision-support system combining Retrieval-Augmented Generation (RAG) with Hybrid Decision-Making. Unlike conventional RAG systems…
- arXiv cs.AIresearchGeoFitBench-LLM: A Benchmark for Assessing Large Language Models' Capabilities in Geospatial Data Requirements Elicitation for Geo-Analytic Tasks
arXiv:2610.12669v1 Announce Type: cross Abstract: Large language models are becoming central to GeoAI systems, yet their evaluation rarely considers the elicitation of geospatial data requirements. This omission is criti…
- arXiv cs.AIresearchUnsupervised Multimodal Reward Modeling
arXiv:2610.12707v1 Announce Type: cross Abstract: Vision-language reward models are trained on annotated preference pairs, which are expensive to collect, can disagree, and are ranked by benchmark accuracy. We build one…
- arXiv cs.AIresearchKITE: A Knowledge-Injected Temporal Encoder for Time-Series Event Forecasting
arXiv:2610.12738v1 Announce Type: cross Abstract: Forecasting rare, high-cost events in multivariate time-series systems is hard because their precursors are seldom observed, so a purely data-driven model has little to l…
- arXiv cs.AIresearchSemantic Views: Certified Reuse of LLM Labels Across Semantic Queries
arXiv:2610.12768v1 Announce Type: cross Abstract: Semantic query engines evaluate natural-language filters with a large language model, and every evaluation is a costly oracle call. Optimizers that reduce these calls fac…
- arXiv cs.AIresearchREDaCT: Collapse-Free Unlearning in Large Reasoning Models
arXiv:2610.12769v1 Announce Type: cross Abstract: Unlearning in large reasoning models (LRMs) requires removing forget-set information from both the reasoning trace and the final answer. Existing approaches achieve this…
- arXiv cs.AIresearchMeta-Reflective Evolution Effectively Improves Production AI Systems
arXiv:2610.12794v1 Announce Type: cross Abstract: Large Language Model (LLM)-guided evolutionary optimization has successfully improved prompts, programs, and even entire AI systems. Most optimization methods select a ca…
- arXiv cs.AIresearchEfficient Audio-Visual Speech Enhancement via Reinforcement Learning-Based Neural Architecture Search
arXiv:2610.12796v1 Announce Type: cross Abstract: Audio-visual speech enhancement (AVSE) improves speech intelligibility in noisy environments but often relies on computationally expensive models unsuitable for resource-…
- arXiv cs.AIresearchLeWAM: End-to-End World and Action Modeling with JEPAs
arXiv:2610.12818v1 Announce Type: cross Abstract: Joint-embedding predictive architectures (JEPAs) learn compact latent world models without pixel reconstruction. Existing JEPA pipelines typically train visual representa…
- arXiv cs.AIresearchCOFFEQA: Counterfactual-Guided Few-Shot Distillation for LLM Question Answering
arXiv:2610.12837v1 Announce Type: cross Abstract: Knowledge distillation aims to train a smaller student model to mimic a larger teacher model for resource-constrained environments. However, the process of distillation i…
- arXiv cs.AIresearchWhen Generative Models Act as Simulators
arXiv:2610.12847v1 Announce Type: cross Abstract: A simulator summarizes an environment in a state that users can read and change. Generative models learn representations of the world without explicit specification, but…
- arXiv cs.AIresearchLocalized Generative Design of Ship Hulls with Diffusion Models and High-Fidelity Assessment
arXiv:2610.12870v1 Announce Type: cross Abstract: Practical ship design requires localized refinement and performance evaluation. This paper presents a reward-directed diffusion framework for localized hull optimization.…
- arXiv cs.AIresearchSurfing the Activation Wave: Analyzing the Information and Geometry of LLM Hidden States for Speculative Decoding
arXiv:2610.12939v1 Announce Type: cross Abstract: Decoder hidden states from different layers of large language models (LLMs) often encode complementary information. However, existing speculative decoding methods typical…
- arXiv cs.AIresearchMotionCoach: Constructing Multimodal Guidance for Video Professional Motion Editing
arXiv:2610.12963v1 Announce Type: cross Abstract: Video editing models achieve strong performance in modifying appearance, backgrounds, styles, and motions, empowered by large-scale training data. Beyond general scenario…
- arXiv cs.AIresearchThe Evidence Anchoring Boundary: Where Visual Interventions Matter and When They Help in Multimodal LLMs
arXiv:2610.12981v1 Announce Type: cross Abstract: Visual interventions provide a training-free way to reduce hallucinations in multimodal large language models. However, an intervention may have little leverage at some l…
- arXiv cs.AIresearchWhat Should SFT Learn? Conditional Information Gain for Generalizable Post-Training
arXiv:2610.12995v1 Announce Type: cross Abstract: Supervised fine-tuning (SFT) remains a widely used approach for post-training large language models, because fixed demonstrations can directly introduce desired reasoning…
- arXiv cs.AIresearchGeneralized Proximal Policy Optimization: Unifying LLM RL Algorithms from the Perspective of Divergence
arXiv:2610.13015v1 Announce Type: cross Abstract: Reinforcement learning has become the foundational paradigm for enhancing the reasoning capabilities of LLMs. Yet the dominant approach, Proximal Policy Optimization (PPO…
- arXiv cs.AIresearchDon't Guess, Just Ask: Evaluating Interactive Requirement Clarification for Repository-Level Software Development
arXiv:2610.13022v1 Announce Type: cross Abstract: As LLMs become increasingly capable of implementing software, a critical bottleneck emerges: whether they truly understand what users want. Users often begin with vague i…
- arXiv cs.AIresearchGeoGR: Spatial-Semantic Token Compression and Group Routing for Efficient 3D VLMs
arXiv:2610.13035v1 Announce Type: cross Abstract: Multi-view representations enable pretrained image and video vision-language models to understand 3D scenes, but their long visual token sequences make inference costly.…
- arXiv cs.AIresearchCoBe: A Benchmark For Conversational Counterfactual Text Editing
arXiv:2610.13057v1 Announce Type: cross Abstract: Counterfactual text editing is the ability to revise a given text in light of a hypothetical change while preserving facts not causally affected by that change. When aske…
- arXiv cs.AIresearchEvolving Creatives with Clicks: CTR-Driven Advertising image Editing via Attribute Disentanglement and Preference Alignment
arXiv:2610.13066v1 Announce Type: cross Abstract: The optimization objective for e-commerce advertising creatives has shifted from aesthetic quality to online business metrics like click-through rate (CTR). Recent method…
- arXiv cs.AIresearchContinually Post-training Small Models via Synthetic Self-distillation
arXiv:2610.13113v1 Announce Type: cross Abstract: Continual post-training faces a dual challenge of acquisition and retention: models must learn new capabilities while preserving prior knowledge. While Self-Distillation…
- arXiv cs.AIresearchDoes Learning Recurrent Networks Matter in Partially Observable RL?
arXiv:2610.13197v1 Announce Type: cross Abstract: Reinforcement learning (RL) under partial observability typically re- lies on recurrent networks to encode temporal information, with the recurrent module trained jointly…
- arXiv cs.AIresearchBeyond Actions: Learning Future Operation Targets for Vision-Language-Action Models
arXiv:2610.13212v2 Announce Type: replace-cross Abstract: Actions specify what a robot should execute, but do not describe what those commands will achieve. Vision-language-action (VLA) policies nevertheless learn primar…
- arXiv cs.AIresearchReSFlow: Rectifying Categorical Flows on Statistical Manifolds for One-Step Generation
arXiv:2610.13225v2 Announce Type: replace-cross Abstract: Generative modeling for discrete data has been significantly advanced by learning continuous vector fields on the statistical manifolds to transport simple priors…
- arXiv cs.AIresearchHow Reliable are AMR Parser Outputs? Uncertainty Estimation for AMR Parsing
arXiv:2610.13236v2 Announce Type: replace-cross Abstract: Recent advances have significantly improved the performance of Abstract Meaning Representation (AMR) parsers, yet determining when their outputs can be trusted re…
- arXiv cs.AIresearchTabular Foundation Models Are Effectively Shallow
arXiv:2610.13253v1 Announce Type: cross Abstract: In-context-learning Tabular Foundation Models (TFMs) are uniquely suited to scenarios with limited resources or privacy constraints, as they solve new tasks without addit…
- arXiv cs.AIresearchAgreement Without Alignment: Label-Free Rubric Clarification Increases LLM-Judge Consensus but Not Agreement with Humans
arXiv:2610.13303v1 Announce Type: cross Abstract: LLM judges are increasingly deployed as panels, and disagreement among panel members is often read as a symptom of an unclear rubric. We test a label-free procedure that…
- arXiv cs.AIresearchIn-Context Learning for System Identification: Probabilistic Formulation, Long-Sequence Handling, and Benchmark Validation
arXiv:2610.13411v1 Announce Type: cross Abstract: Traditional system identification typically involves two steps: initially, training data is processed to estimate a model of the system of interest; then, the model is us…
- arXiv cs.AIresearchCNN-Transformer with Sigmoid Label for High-Resolution Sub-Degree DOA Estimation
arXiv:2610.13433v1 Announce Type: cross Abstract: This paper addresses the grid quantization error of deep-learning (DL)-based direction-of-arrival (DOA) estimation. Classification- based DOA methods discretize the conti…
- arXiv cs.AIresearchWhen MLLMs miss what humans read: diagnosing and learning to read squint text
arXiv:2610.13465v1 Announce Type: cross Abstract: Multimodal LLMs can fail to read text embedded in an image's coarse structure (text a human recovers by squinting) and report that no text is present. We name this class…
- arXiv cs.AIresearchBeyond Temporal Proximity: Explanatory World Models for Credit Assignment in Reinforcement Learning
arXiv:2610.13504v1 Announce Type: cross Abstract: Generative Recommenders increasingly use Reinforcement Learning (RL) to optimize outcomes that may appear hundreds or thousands of interactions after a decision. Standard…
- arXiv cs.AIresearchRESCUE: Is This Alarm Worth Acting On? Counterfactual Evaluation of VLA Runtime Monitors
arXiv:2610.13516v1 Announce Type: cross Abstract: Runtime failure monitors for robot policies are typically evaluated by how accurately they distinguish failed from successful executions. Yet an accurate alarm may arrive…
- arXiv cs.AIresearchDefence against the dark arts of distillation: Unsupervised discovery of subliminal learning
arXiv:2610.13536v1 Announce Type: cross Abstract: When a language model is distilled from a teacher that shares its base parameters, the student can inherit behavioural traits of the teacher---including misalignment---ev…
- arXiv cs.AIresearchAudioTETRIS: Token Elimination via Transitional Importance Signals for Efficient Audio LLMs
arXiv:2610.13600v1 Announce Type: cross Abstract: Large Audio Language Models (LALMs) achieve strong performance on speech-text tasks, but standard audio encoders produce hundreds of tokens for even short audio clips. Be…
- arXiv cs.AIresearchSWE-DeceptionBench: Benchmarking Deceptive Behaviors of LLMs Across the Software Development Lifecycle
arXiv:2610.13626v1 Announce Type: cross Abstract: Large Language Models (LLMs) are increasingly used throughout the software development lifecycle. This growing reliance on LLMs raises concerns about their trustworthines…
- arXiv cs.AIresearchModel-agnostic Knowledge Augmented Machine Learning for Weakly Annotated Bioimaging Data
arXiv:2610.13637v1 Announce Type: cross Abstract: Machine learning models are increasingly used for automated feature extraction from weakly annotated bioimaging data, such as microscopy and cytomorphology images. Due to…
- arXiv cs.AIresearchSelectArena: A Modern Benchmark for Feature Selection on Real-world Tabular Data
arXiv:2610.13644v2 Announce Type: replace-cross Abstract: Selecting a subset of features (columns) from a tabular dataset is a ubiquitous challenge in real-world applications. Industry experts seek to minimize model trai…
- arXiv cs.AIresearchPretraining and Test-Time Training Budgets Matter for ARC: A Controlled Study of VARC and TRM
arXiv:2610.13660v1 Announce Type: cross Abstract: The Abstraction and Reasoning Corpus (ARC) is increasingly used to evaluate whether neural networks can infer task-specific transformations from few demonstrations. Recen…
- arXiv cs.AIresearchInject a Turn, Not a Token: Learning Where to Redirect LLM Reasoning for Self-Correction
arXiv:2610.13672v1 Announce Type: cross Abstract: Large language models struggle to reliably identify and correct errors in their own reasoning. Unlike prior training-based approaches that fine-tune an LLM to generate re…
- arXiv cs.AIresearchNot All Dimensions Are Equal: Outlier-Aware Attention Linearization with Grouped Quadratic Kernels
arXiv:2610.13674v1 Announce Type: cross Abstract: Linearizing softmax attention offers a route to efficient Transformer inference, but preserving pretrained attention remains challenging. We identify an important source…
- arXiv cs.AIresearchVariance-Reduced Policy Gradient is Nearly Minimax-Optimal
arXiv:2610.13692v1 Announce Type: cross Abstract: Standard Policy Gradient (PG) algorithms suffer from high estimation variance because they rely on Monte Carlo updates. While off-policy variance reduction mitigates this…
- arXiv cs.AIresearchCalibration in One Go: Jointly Learning Accurate and Calibrated Neural Networks
arXiv:2610.13693v1 Announce Type: cross Abstract: Neural networks (NNs), while highly flexible and versatile, are often poorly calibrated. Post-hoc calibration methods adjust the outputs of an already trained model using…
- arXiv cs.AIresearchAmortized Counterfactual Prediction with Graph Neural Networks
arXiv:2610.13724v1 Announce Type: cross Abstract: Counterfactual inference asks how an observed outcome might have differed under different circumstances, such as whether a patient would have recovered had they received…
- arXiv cs.AIresearchLAPT-Scale: Scaling Embodied Policies with Latent Action Pre-Training on Human Videos
arXiv:2610.13732v1 Announce Type: cross Abstract: Scaling generalist robotic policies is limited by the high cost of collecting embodied interaction data. Large-scale egocentric human videos offer a rich alternative, but…
- arXiv cs.AIresearchOrthros: Phase-Aware Heterogeneous Attention for Efficient Transformers
arXiv:2610.13735v1 Announce Type: cross Abstract: Efficient attention mechanisms, such as linear and sliding window attention, aim to reduce the computational complexity of Transformers from quadratic to linear. However,…