AI news archive.
19,861 articles, newest first, from the outlets listed on the sources page.
- arXiv cs.AIresearchCross-Lingual Transfer of Evil Vectors Is Directionally Asymmetric
arXiv:2610.17880v1 Announce Type: cross Abstract: Persona vectors are linear directions in the residual-stream activations of LLMs that encode behavioral traits and enable fine-grained control over model behavior. Among…
- arXiv cs.AIagentsInteraction-as-a-Judge
arXiv:2610.17895v3 Announce Type: replace-cross Abstract: Coding agents are tasked with producing sophisticated interactive applications such as websites, desktop applications, and games. While humans assess these applic…
- arXiv cs.AIresearchEvaluating the Causal Faithfulness of SHAP as a Candidate Attribution Method for SAE Features
arXiv:2610.17898v1 Announce Type: cross Abstract: Sparse autoencoders decompose language model representations into interpretable features, but feature interpretation in this literature remains predominantly correlationa…
- arXiv cs.AIresearchVisual Dependence Does Not Imply Answerability: A Systematic Analysis of Answerability Signals in Frozen InstructBLIP
arXiv:2610.17902v1 Announce Type: cross Abstract: Visual question answering models often produce plausible responses even when the available visual evidence is insufficient to support a valid answer. Although prior work…
- arXiv cs.AIagentsDriveWLA: Unifying Future Dynamics and Scene Semantics through World-Language-Action Modeling for Autonomous Driving
arXiv:2610.17903v1 Announce Type: cross Abstract: Vision-Language Models (VLMs) and world models have demonstrated strong capabilities in semantic understanding and future scene evolution for autonomous driving. However,…
- arXiv cs.AIresearchSpecCloak: Probing Harmful Code Generation through Engineering Specifications
arXiv:2610.17907v1 Announce Type: cross Abstract: The code generation capabilities of large language models (LLMs) can equally be misused to implement harmful functionality, underscoring the need to understand their safe…
- arXiv cs.AIresearchLinMorph: Learning to Morph Weights for Efficient Video DiT Linearization
arXiv:2610.17914v1 Announce Type: cross Abstract: Post-hoc linearization is an attractive way to accelerate pretrained Video Diffusion Transformers (Video DiTs), yet existing methods largely treat the quality loss induce…
- arXiv cs.AIresearchCineMem: Attention-Aligned Fine-grained Memory Retrieval for Multi-Shot Long Video Generation
arXiv:2610.17926v1 Announce Type: cross Abstract: Autoregressive video generation relies on memory retrieval to model long-range dependencies. This capability becomes particularly critical for multi-shot generation, wher…
- arXiv cs.AIagentsGLST: Global-Local Spatio-Temporal Memory Decomposition Framework for Online Chained-Goal Navigation
arXiv:2610.17927v1 Announce Type: cross Abstract: Autonomous navigation in real-world environments often requires long-horizon operation under partial observability and evolving goals. Existing mapless navigation approac…
- arXiv cs.AIresearchMERIT: Multimodal Evidence Reasoning over Image Tables
arXiv:2610.17937v1 Announce Type: cross Abstract: Table reasoning with language models typically assumes access to clean textual tables. In practice, tables are often serialized into text, which can corrupt cell contents…
- arXiv cs.AIresearchAI-Enhanced Hyperspectral Imaging for Intraoperative Detection of Low-Grade Glioma: A Clinical Study
arXiv:2610.17944v1 Announce Type: cross Abstract: Current image-guided neuronavigation techniques, such as intraoperative MRI and ultrasound, are limited by high costs, limited sensitivity and operational complexity. The…
- arXiv cs.AIresearchHaloBlock: Faster and Better Decoding for Diffusion Large Language Models
arXiv:2610.17948v1 Announce Type: cross Abstract: Diffusion large language models (dLLMs) have attracted increasing research attention due to their bidirectional attention and arbitrary order decoding. However, there are…
- arXiv cs.AIresearchFETALFusion: Resolution-Aware State-Space Modeling for Multi-Domain Fetal Head Segmentation
arXiv:2610.17951v1 Announce Type: cross Abstract: Head circumference (HC) measured from 2D ultrasound is a routine marker of fetal growth, and automating it needs segmentations whose boundaries hold up on individual scan…
- arXiv cs.AIresearchVirtual Head Attention
arXiv:2610.17959v1 Announce Type: cross Abstract: The design of attention mechanisms in long-context large language models faces an ``impossible triangle'' among KV cache size, training and prefill FLOPs, and model quali…
- arXiv cs.AIresearchBetter Capturing Interactions Between Products in~Retail: Revisited Negative Sampling for~Basket Choice Modeling
arXiv:2610.17963v1 Announce Type: cross Abstract: Brick-and-mortar retailers face many different challenges that involve understanding thoroughly its products catalog and customer preferences. In particular, assortment o…
- arXiv cs.AIresearchBELT : BEhavior-Aligned Low-Rank KV Cache Compression with Learned Top-k Rank Allocation
arXiv:2610.17966v1 Announce Type: cross Abstract: Long-context inference makes the key-value (KV) cache a major memory bottleneck for large language models. We propose BELT, a low-rank KV cache compression framework. Its…
- arXiv cs.AIresearchSeeing Through the Forgery Is Not Enough: Regressive Sycophancy under Fabricated Evidence in Medical Vision-Language Models
arXiv:2610.17989v1 Announce Type: cross Abstract: A medical vision-language model may analyze a given CT scan and correctly detect both kidneys in view. Continuing this conversation, if a user provides a prescription wit…
- arXiv cs.AIresearchSaliency-Guided Replay for Continual Learning in Deep Spiking Neural Networks
arXiv:2610.17998v1 Announce Type: cross Abstract: Spiking Neural Networks (SNNs) are promising for energy-efficient neuromorphic computing, but they still suffer from catastrophic forgetting when learning tasks sequentia…
- arXiv cs.AIresearchMatch the Canvas, Not Just the Answer: Predictability-Aware SFT for Diffusion Language Models
arXiv:2610.18001v1 Announce Type: cross Abstract: Diffusion language models (DLMs) enable flexible parallel generation by resolving tokens according to the model's evolving predictability rather than a fixed left-to-righ…
- arXiv cs.AIagentsSWEPatchTest: Evaluating Software Tests Beyond Reproduction
arXiv:2610.18004v1 Announce Type: cross Abstract: Coding agents use generated tests to decide whether a proposed patch satisfies an issue's requirements. A suite that fails on the buggy program and passes after applying…
- arXiv cs.AIagentsCost-Efficient Privacy Enforcement for Email-Generating LLM Agents under Strategic Information Requests
arXiv:2610.18007v1 Announce Type: cross Abstract: LLM agents now act as delegated communication intermediaries: they read private context, draft messages, and send externally visible emails on behalf of a user. Existing…
- arXiv cs.AIresearchCorrecting or Continuing? What On-Policy Distillation Learns from Student Errors
arXiv:2610.18011v1 Announce Type: cross Abstract: On-policy distillation (OPD) trains a student on its own rollouts under a frozen teacher, mitigating the distribution mismatch of off-policy supervision. While prior work…
- arXiv cs.AIagentsLatentRemedy: Efficient Remedy For Non-Prefix KV Cache Reuse in Multi-Agent Systems
arXiv:2610.18014v1 Announce Type: cross Abstract: Non-prefix KV cache reuse is a promising approach for improving inference efficiency in multi-agent systems as agents share the same context blocks. However, when agents…
- arXiv cs.AIresearchChampionKV: Head-Wise Adaptive Budget Allocation for KV Cache Compression
arXiv:2610.18026v1 Announce Type: cross Abstract: Some observation-window-based KV cache compressors use prompt attention to identify important tokens but allocate uniform or predefined budgets across heads, overlooking…
- arXiv cs.AIresearchCan Machine Learning Potentials Be More Accurate Than the Methods They Are Trained On?
arXiv:2610.18027v1 Announce Type: cross Abstract: Although machine learning potentials (MLPs) generalize across configurational and chemical compound spaces, their accuracy is commonly believed to be limited by their tra…
- arXiv cs.AIresearchScaling Vision Transformers Without Training: Prune-and-Merge Token Compression
arXiv:2610.18039v2 Announce Type: replace-cross Abstract: Vision Transformers achieve strong performance but incur high computational costs from quadratic token interactions. Training-free token reduction methods improve…
- arXiv cs.AIresearchHigh-fidelity Three-Dimensional Flow Field Reconstruction Driven by Turbulent Fluid Dynamics
arXiv:2610.18045v1 Announce Type: cross Abstract: High-fidelity three-dimensional turbulent flow fields are essential for scientific and engineering analyses, yet obtaining them through direct numerical simulation or hig…
- arXiv cs.AIresearchinLoRA-MINT: Internal Representations for Auditing Training Data in Domain-adapted LLMs
arXiv:2610.18051v1 Announce Type: cross Abstract: This work investigates the application of the Membership Inference Test (MINT) to large language models (LLMs) fine-tuned with LoRA, aiming to determine whether specific…
- arXiv cs.AIresearchFrom Policy Evaluation to Off-Policy Control: Provable Emergence of In-Context Q-Learning
arXiv:2610.18057v1 Announce Type: cross Abstract: Transformers can implement reinforcement-learning algorithms in context, yet how off-policy control emerges within Transformers remains poorly understood. We study Q-lear…
- arXiv cs.AIresearchUncertainty Evaluation: Bridging Machine Learning and Measurement Science
arXiv:2610.18066v1 Announce Type: cross Abstract: Reliable uncertainty evaluation is essential for assessing the trustworthiness of machine learning (ML) predictions, particularly in safety-critical applications. This pa…
- arXiv cs.AIagentsAPT: Feedback-Driven Agentic Post-Training for Text-to-Image Generation
arXiv:2610.18097v1 Announce Type: cross Abstract: Post-training is crucial for improving text-to-image models, yet existing pipelines typically rely on manually curated training data and fixed training distributions. We…
- arXiv cs.AIresearchA CaLLM Judge is a Better Judge: Mitigating Bias in LLM-as-a-judge
arXiv:2610.18108v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly used as judges to predict which of two candidate answers a human evaluator would prefer, providing a scalable alternative wh…
- arXiv cs.AIresearchWhen Missingness Follows the Graph: Beyond Feature Propagation for Missing Features
arXiv:2610.18113v1 Announce Type: cross Abstract: Graph Neural Networks (GNNs) commonly assume that node features are fully observed, despite missing features being pervasive in real-world applications. By leveraging the…
- arXiv cs.AIresearchCertifying Almost-Sure Reachability in Unknown MDPs
arXiv:2610.18119v1 Announce Type: cross Abstract: {\em Specification-Guided Reinforcement Learning (RL)} has largely concerned itself with \emph{learning} an optimal policy against a formal specification in an unknown {\…
- arXiv cs.AIresearchH-DDCM: Hadamard Codebooks for Fast Generative Image and Video Compression
arXiv:2610.18125v1 Announce Type: cross Abstract: Training-free diffusion-based compression methods repurpose the generative priors of a diffusion or a flow model to compress data without retraining them as task-specific…
- arXiv cs.AIresearchWhen Do Learned Blocks Preserve Causal Meaning? Block Identifiability and Constructive Causal Abstractions
arXiv:2610.18155v2 Announce Type: replace-cross Abstract: Standard identifiability theory assumes one-dimensional latent factors, but learned representations can encode latent information across multidimensional blocks w…
- arXiv cs.AIagentsFan Out or Relay? Ownership-Aware Routing for LLM Agents Querying Siloed Tables
arXiv:2610.18157v1 Announce Type: cross Abstract: Inside organizations, a question's data often sit with several owners, who query their own tables but never hand them over. An LLM coordinator can fan out, sending the qu…
- arXiv cs.AIresearchClaim Framing as a Driver of LLM Sycophancy
arXiv:2610.18159v1 Announce Type: cross Abstract: Sycophancy in Large Language Models, the tendency to align responses with user preferences over ground truth, has largely been studied using questions without a clear gro…
- arXiv cs.AIresearchLoOPSD: On-Policy Self-Distillation with Looped Teachers
arXiv:2610.18162v1 Announce Type: cross Abstract: On-policy self-distillation improves a model using guidance from the model itself, but it still requires a teacher that provides more useful predictions than the student.…
- arXiv cs.AIresearchLifting Finite-Sample Gaussian Approximation of Wide Neural Networks to Functional Convergence
arXiv:2610.18166v1 Announce Type: cross Abstract: Randomly initialised neural networks of width $m$ converge to Gaussian processes as $m \to \infty$, with two types of quantitative bounds characterising this convergence.…
- arXiv cs.AIresearchLooped Reasoning Finishes Earlier Than You Think
arXiv:2610.18174v1 Announce Type: cross Abstract: Looped reasoning with dynamic depth can be viewed as relaxation toward solution attractors in a learned dynamical landscape. In practice, however, latent trajectories oft…
- arXiv cs.AIresearchA Few More Tokens Ahead: Reinforcing Parallelism in Diffusion LLMs
arXiv:2610.18200v1 Announce Type: cross Abstract: Parallel generation is widely regarded as an inherent advantage of diffusion large language models. Yet our fixed-quota evaluation of four representative models reveals s…
- arXiv cs.AIresearchHow Reliable Are Large Language Models for Soil Organic Carbon Assessment? An Australian Benchmark
arXiv:2610.18201v1 Announce Type: cross Abstract: Soil organic carbon (SOC) is an indicator of agricultural soil condition and carbon storage, but numerical outputs from large language models (LLMs) require validation be…
- arXiv cs.AIresearchReset Early, Reset Often, Eliminate Models
arXiv:2610.18204v1 Announce Type: cross Abstract: Connection-tableau provers typically construct closed tableaux by backtracking search. This gives lightweight proof procedures, but their low-memory search state leaves l…
- arXiv cs.AIagentsTextualRL: Textual Reinforcement Learning for Context Management
arXiv:2610.18216v1 Announce Type: cross Abstract: LLM agents can adapt to task feedback by revising the textual context that guides their behavior, without changing their model parameters. However, turning execution rewa…
- arXiv cs.AIresearchA Controlled Evaluation of Curvature Inspired Scalar Feature Scaling for Domain Generalization
arXiv:2610.18227v1 Announce Type: cross Abstract: Domain-specific curvature is an appealing idea for domain generalization (DG), but curvature learning can denote substantially different computations. This paper evaluate…
- arXiv cs.AIagentsHolding Agents Accountable with Verifiable Agent Insurance
arXiv:2610.18230v1 Announce Type: cross Abstract: While AI agents are useful for many tasks, they regularly misbehave in destructive ways, such as deleting important data. While prior work has focused on preventing such…
- arXiv cs.AIresearchHeading Hierarchy Reconstruction for Complex and Long Documents with Level-Wise Multimodal Memory
arXiv:2610.18238v1 Announce Type: cross Abstract: Vision-language models (VLMs) have been widely adopted for document parsing and optical character recognition (OCR), as they can accurately extract page-level elements. H…
- arXiv cs.AIresearchLiFUSE: Frequency-Spatial Unified Self-Supervised Pretraining for Liver Disease Analysis on Non-contrast CT
arXiv:2610.18243v1 Announce Type: cross Abstract: Given the growing global burden of liver tumors and diffuse liver diseases, non-contrast computed tomography (NCCT) is increasingly used for opportunistic screening becau…
- arXiv cs.AIresearchFALLNEST: Dual-Domain Dataset Condensation of Imbalanced IMU Data for Wearable Fall Detection
arXiv:2610.18260v1 Announce Type: cross Abstract: Wearable fall detection is hampered by rare fall events, scarce labeled IMU data, and the cost of training deep models on large, imbalanced corpora. Existing methods eith…