AI news archive.
19,861 articles, newest first, from the outlets listed on the sources page.
- arXiv cs.AIresearchA Tropical Geometry View of Forgetting: A Per-Unit Projector for Knowledge-Preserving Fine-Tuning
arXiv:2610.04670v1 Announce Type: new Abstract: Fine-tuning a language model on new text degrades what it already does. Replay-free projectors such as Adam-NSCL and GPM forbid one shared subspace of a layer's inputs in e…
- arXiv cs.AIagentsMASBench: Benchmarking LLM-based Multi-Agent Collaboration under Partial Observability
arXiv:2610.04672v1 Announce Type: new Abstract: Large language models (LLMs) have progressively evolved into the core of autonomous agents. Building on this progress, LLM-based multi-agent systems (MAS) coordinate multip…
- arXiv cs.AIagentsWhen Debate Helps: Proposal Supply and Verification-Aware Readout in Multi-Agent Reasoning
arXiv:2610.04686v1 Announce Type: new Abstract: Multi-agent debate can improve reasoning, yet often fails to beat simple majority voting. We argue that successful debate requires two distinct mechanisms: proposal supply…
- arXiv cs.AIresearchRAGStress: A controlled benchmark for evaluating retrieval-augmented generation under knowledge-base degradation
arXiv:2610.04691v1 Announce Type: new Abstract: Retrieval-Augmented Generation (RAG) is typically evaluated under the implicit assumption that the underlying knowledge base (KB) is clean, leaving the behaviour of RAG sys…
- arXiv cs.AIagentsPenumbra: Sample-Efficient Adversarial Search for Regulatory Obligations
arXiv:2610.04693v3 Announce Type: replace Abstract: Agents are entering finance, healthcare and law, sectors where a violation leaves no lexical signature and carries real penalties. Whether an omission is material, or a…
- arXiv cs.AIagentsNot Self-Decidable: LLMs Cannot Draw the Boundary of What an Agent Verifier Can Check
arXiv:2610.04699v2 Announce Type: replace Abstract: A verifier for an agent faces rules of two kinds: the ones a fixed check can settle and the ones that require a judge. A team that derives its own checks fixes that spl…
- arXiv cs.AIagentsLearning to Clarify Underspecified Intents Under Limited Interaction
arXiv:2610.04719v1 Announce Type: new Abstract: AI assistants receive requests that leave out information needed for a good outcome, for example about users' preferences or goals. They must then either speculate or ask f…
- arXiv cs.AIresearchPyINE: A Framework for Scalable Elicitation and Oversight via Code Execution
arXiv:2610.04737v2 Announce Type: replace Abstract: Reasoning models can remain capable of solving a task while still defaulting to cheaper but misleading shortcuts. This creates a central oversight problem: when a model…
- arXiv cs.AIagentsToward a Locally Deployable Agentic Co-Scientist: Small-Model Planning for Early-Stage Drug Discovery
arXiv:2610.04740v1 Announce Type: new Abstract: Early-stage computational drug discovery requires coordinating heterogeneous scientific tools across multi-step workflows. We present a lightweight, tool-augmented framewor…
- arXiv cs.AIresearchFormalizing the Moral Evaluation of Speech Acts: Truthfulness, Lies and Ethical Dilemmas
arXiv:2610.04747v1 Announce Type: new Abstract: In life-or-death situations, a benevolent lie may appear more moral than telling the truth. Yet such lies can backfire, producing unintended and sometimes fatal consequence…
- arXiv cs.AIagentsAgentic discovery of blood biomarker from distilled private health records
arXiv:2610.04749v1 Announce Type: new Abstract: Routine complete blood counts (CBCs) could yield new biomarkers, but the private records needed to evaluate candidates cannot be shared with frontier language model agents…
- arXiv cs.AIresearchDynamic Routing as a New Dimension for Test-time Versatility of LLMs
arXiv:2610.04751v1 Announce Type: new Abstract: Beyond scaling their parameters and data, large language models currently gain versatility on new problems along a single axis: the tokens they spend on chain-of-thought (C…
- arXiv cs.AIresearchWhen Is Enough Enough in Self-Evolving LLM Systems?
arXiv:2610.04756v1 Announce Type: new Abstract: Self-evolving large language model (LLM) systems repeatedly propose, evaluate, and incorporate updates to prompts, skills, or other persistent artifacts. Despite their grow…
- arXiv cs.AIresearchNot All Answers Are Contextually Persuadable: Inference Dynamics in Large Language Models under Contextual Influence
arXiv:2610.04791v1 Announce Type: new Abstract: At the core of modern prompting techniques is contextual sensitivity, the ability of large language models to adapt their predictions based on inference-time context. Despi…
- arXiv cs.AIresearchDo More Modalities Always Help? A Geometric Perspective on Missing-Modality Robustness
arXiv:2610.04792v2 Announce Type: replace Abstract: Missing modality remains a longstanding challenge in multimodal learning. Existing methods typically address this issue through modality recovery or adaptive strategies…
- arXiv cs.AIagentsPressure, Context, and Machine Self-Control: A Criminological Test of Reward Hacking in Generative AI Models
arXiv:2610.04793v1 Announce Type: new Abstract: Recent incidents show that AI agents sometimes reach measured goals through unsanctioned means. This study applies self-control, general strain, anomie, neutralization and…
- arXiv cs.AIagentsKnowing the Store: What a Memory Backend Must Write Down Before an Agent Can Read It
arXiv:2610.04794v1 Announce Type: new Abstract: An agent with long-term memory can answer from a record it should no longer use, such as a plan the user later cancelled. We ask what a memory store must expose for an agen…
- arXiv cs.AIagentsMedImageOSWorld: Benchmarking GUI Agents for Medical Image Consoles
arXiv:2610.04800v1 Announce Type: new Abstract: Graphical consoles offer a practical interface for medical acquisition assistance, allowing agents to work through the controls and visual feedback used by human operators.…
- arXiv cs.AIagentsAgent Behavior as Code: Efficient and Robust LLM Agents with Programmatic Specifications
arXiv:2610.04824v1 Announce Type: new Abstract: AI agents based on foundation models (FMs) have demonstrated strong capabilities to perform complex open-ended tasks. However, they face some common challenges in practice:…
- arXiv cs.AIresearchDICE: Decoupling Capability from Intervention Necessity in LLM Tutoring
arXiv:2610.04825v1 Announce Type: new Abstract: Fluent guidance is not the same as useful intervention. LLM tutors are typically trained to generate the next teacher utterance, implicitly assuming that every student turn…
- arXiv cs.AIagentsMemTrace: State-Consistent Memory for Long-Horizon Coding Agents
arXiv:2610.04838v3 Announce Type: replace Abstract: As coding agents take on long-horizon software evolution tasks spanning multiple files and stages, longer execution trajectories introduce two coupled challenges: (1) a…
- arXiv cs.AIresearchGitSwarm: Decentralized Compounding Inference
arXiv:2610.04862v1 Announce Type: new Abstract: Long-horizon problem solving and scientific research require computation to accumulate across successive attempts. Partial solutions, experimental findings, and unsuccessfu…
- arXiv cs.AIagentsFrom Memory to Guide: Spatio-Temporal Composer for Procedural Coding Memory
arXiv:2610.04868v1 Announce Type: new Abstract: Memory-augmented agents typically integrate procedural knowledge by injecting retrieved skills directly into text prompts. This approach dangerously equates readable text w…
- arXiv cs.AIresearchSpecFold: Folding Multi-Branch Redundancy for Faster Speculative Decoding in Diffusion Language Models
arXiv:2610.04875v2 Announce Type: replace Abstract: Diffusion large language models (DLLMs) generate text through iterative block denoising, and multi-branch speculative decoding accelerates this process by verifying a m…
- arXiv cs.AIagentsForkPilot: Self-Evolving Policy for Retrospective Search in Long-Horizon Agents
arXiv:2610.04889v1 Announce Type: new Abstract: Interactive language-model agents increasingly solve complex tasks through long-horizon, multi-call reasoning, where errors in beliefs or actions can compound across tool i…
- arXiv cs.AIagentsSelf-Evaluating Recursive Agents
arXiv:2610.04902v1 Announce Type: new Abstract: Recursive language-model agents decompose tasks and delegate subtasks to child instances of the same policy, forming a tree of work. Training them, however, is hard: the fi…
- arXiv cs.AIagentsVideoResearchAgent: Grounded Task Synthesis and Sim-to-Real RL for Open-Web Video Research
arXiv:2610.04911v1 Announce Type: new Abstract: Existing deep research agents are designed primarily for text- and image-based web sources, while video reasoning systems typically assume that relevant videos are provided…
- arXiv cs.AIresearchMonitorability Disposition in Large Reasoning Models
arXiv:2610.04914v1 Announce Type: new Abstract: Monitoring the chain-of-thought (CoT) of large reasoning models (LRMs) is a common way to detect misbehavior in real-world practice. However, current monitoring is passive:…
- arXiv cs.AIagentsAre We Measuring Scientific Intelligence? Rethinking the Evaluation of AI Scientists
arXiv:2610.04915v1 Announce Type: new Abstract: AI agents can now carry out data-driven scientific analyses end to end, and benchmarks assess them by giving an agent a question and a dataset and scoring its final answer…
- arXiv cs.AIresearchZero-Shot Time-Series Question Answering via Decoupled Perception and Reasoning
arXiv:2610.04942v1 Announce Type: new Abstract: Time-series question answering (TSQA) requires grounding linguistic queries and diverse answer formats in complex numerical observations. However, existing methods heavily…
- arXiv cs.AIagentsBACAM: Behavior-Aware Continual Agent Merging for Multi-Turn Interaction
arXiv:2610.04966v1 Announce Type: new Abstract: Model merging offers a way to integrate the capabilities of specialized experts, but existing agent merging methods typically require all of them to be available at once. W…
- arXiv cs.AIagentsRuntime Authorization of Self-Generated Subgoals in Long-Horizon Tool-Using AI Agents
arXiv:2610.04975v1 Announce Type: new Abstract: Long-horizon tool-using AI agents create subgoals, replan, delegate work, and compose sibling results. Per-tool permission checks cannot establish that a changing goal grap…
- arXiv cs.AIagentsDo LLMs Understand Sequential Structure? A Controlled Study of Inference and Generation
arXiv:2610.04977v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used as interactive agents and simulators, yet it remains unclear whether they can recover latent sequential structure beyond…
- arXiv cs.AIagentsCIPO: Counterfactual Imagination Policy Optimization for Adaptive Tool Granularity Selection
arXiv:2610.04991v1 Announce Type: new Abstract: Large language model (LLM) agents solve complex tasks through multi-step interactions with external tools. These interactions often contain recurring local tool sequences.…
- arXiv cs.AIresearchEmoRSS: Mitigating Emotion-Induced Over-Refusal in Large Language Models
arXiv:2610.04998v1 Announce Type: new Abstract: Emotional expression can influence the safety decisions of large language models (LLMs), offering a potential avenue for improving safety alignment. Existing studies have m…
- arXiv cs.AIagentsHow corner is a corner case? Percentile control for highway scenario generation
arXiv:2610.05003v1 Announce Type: new Abstract: Generating corner-case scenarios with appropriate adversity in a simulation environment is critical for testing an autonomous vehicle (AV) software stack's safety performan…
- arXiv cs.AIagentsEVISKILL: Grounding Skill Evolution in Replayable Evidence
arXiv:2610.05030v1 Announce Type: new Abstract: Continual skill evolution enables LLM agents to accumulate and refine reusable procedural knowledge from interaction experience without updating model parameters. Its effec…
- arXiv cs.AIresearchCI-JEPA: A Counterfactual Analysis of Latent Representations in Joint-Embedding Predictive Architectures for Self-Supervised Learning
arXiv:2610.05043v1 Announce Type: new Abstract: Self-supervised visual representation learning learns useful features without manual annotations during representation training. The image-based joint-embedding predictive…
- arXiv cs.AIresearchWhy, Where, How: Taxonomy-guided Error Grounding for Code Repair in NL2SQL
arXiv:2610.05060v1 Announce Type: new Abstract: SQL queries that large language models write from natural language questions can execute successfully yet produce incorrect results, so execution alone does not reveal what…
- arXiv cs.AIresearchSharpDraft: Accelerating Long-Context Speculative Decoding with Cardinality-Aware Query Scaling
arXiv:2610.05106v1 Announce Type: new Abstract: Long-form reasoning makes inference expensive, and speculative decoding mitigates this cost by verifying multiple draft tokens in parallel. Its speedup, however, can fade a…
- arXiv cs.AIagentsLexiHorizon: Stabilizing Reinforcement Learning for Long-Horizon Deep Search
arXiv:2610.05119v1 Announce Type: new Abstract: Deep search agents tackle complex knowledge tasks through iterative retrieval, multi-hop reasoning, and evidence synthesis across multiple sources. Existing approaches typi…
- arXiv cs.AIagentsMemory Canonicalization: A Framework and Benchmark for Cross-Model Drift in Persistent LLM Memory
arXiv:2610.05124v1 Announce Type: new Abstract: Persistent memory for Large Language Models (LLMs) has matured rapidly: systems such as MemGPT/Letta, Mem0, and Zep now provide agents with tiered, temporally-aware, model-…
- arXiv cs.AIagentsAutoSciBench: Autonomous Benchmark Generation for Evaluating Scientific Agents
arXiv:2610.05140v1 Announce Type: new Abstract: As agents rapidly evolve, existing benchmarks can become saturated, limiting their ability to distinguish capabilities and reveal remaining failure modes. Particularly in s…
- arXiv cs.AIresearchR1A-PC: Physics-Guided Electromagnetic Inversion of Three-Dimensional Human Point Clouds in Complex Static Environments
arXiv:2610.05144v1 Announce Type: new Abstract: Recovering three-dimensional human geometry from electromagnetic measure?ments in a complex static environment is difficult because strong multipath responses from walls, f…
- arXiv cs.AIresearchBeyond Instruction Following: Learning Grounded Skill-Following with Skill Contracts
arXiv:2610.05161v1 Announce Type: new Abstract: Instruction following typically enforces discrete, response-level requirements, whereas an expert-authored skill prescribes procedural requirements spanning multiple phases…
- arXiv cs.AIagentsMemadapter: Counterfactual Adaptation Against Memory-induced Sycophancy
arXiv:2610.05162v1 Announce Type: new Abstract: Long-term memory enables LLM-based agents to retain and reuse information across tasks and sessions, supporting personalization and long-horizon interactions. However, pers…
- arXiv cs.AIresearchA Safe Action Is Not Enough: Feasible-Future Decoding for Vision-Language-Action Policies
arXiv:2610.05166v2 Announce Type: replace Abstract: A safe action is not necessarily a viable one. A frozen vision-language-action (VLA) policy can favor a locally admissible move that leaves no policy-supported route to…
- arXiv cs.AIresearchCreativeFlow: A One-to-Many Analogical Relation Transfer Method for 3D Asset Generation
arXiv:2610.05167v1 Announce Type: new Abstract: Inspired by cognitive science, we present CREATIVEFLOW, an analogical generation framework that explicitly models analogical divergent thinking to mitigate creative homogen…
- arXiv cs.AIagentsAECG: Asymmetric Experience Consolidation and Governance In Multi-Agent Systems
arXiv:2610.05176v1 Announce Type: new Abstract: Large language model (LLM)-based multi-agent systems increasingly rely on memory to transform execution trajectories into reusable procedural knowledge. Yet repeated retrie…
- arXiv cs.AIagentsFrom Scientific Observations to Mechanisms: Benchmarking Hypothesis Generation by AI Scientists
arXiv:2610.05197v1 Announce Type: new Abstract: Data-driven mechanistic hypotheses are essential to scientific discovery because they explain how underlying processes produce observed phenomena. AI agents and AI scientis…