AI news archive.
19,861 articles, newest first, from the outlets listed on the sources page.
- arXiv cs.AIresearchTopoGraphRAG-Bench: Evaluating Multimodal GraphRAG on Layout-Grounded Evidence Reasoning
arXiv:2610.09360v1 Announce Type: new Abstract: Real-world documents distribute evidence across text, tables, figures, and captions within complex page layouts. Answering complex questions over such documents therefore r…
- arXiv cs.AIagentsDUDA-Bench: Benchmarking LLM Agents on Multimodal Data-Driven Urban Diagnosis
arXiv:2610.09374v1 Announce Type: new Abstract: Urban diagnosis integrates heterogeneous observations to identify urban problems, localize affected areas, and investigate contributing factors, informing evidence-based ur…
- arXiv cs.AIagentsFrom Plausible Hierarchies to Useful Taxonomies: Evaluating Agentic Harnesses on Customer Feedback
arXiv:2610.09377v1 Announce Type: new Abstract: Taxonomies are the symbolic representations through which AI systems organize evidence, aggregate patterns, and answer questions over large document collections. Over custo…
- arXiv cs.AIresearchLet the Library Speak: Self-Advertised Method Selection for Formal Proving
arXiv:2610.09401v1 Announce Type: new Abstract: LLM-based formal provers can retrieve relevant lemmas and prior proofs, but relevance alone does not say whether a mathematical method can be used on the current theorem. A…
- arXiv cs.AIresearchEfficient Reasoning with Flow Language Models
arXiv:2610.09416v1 Announce Type: new Abstract: Flow Language Models (FLMs) have emerged as a continuous-state alternative to discrete diffusion language models, yet the role of their continuous representations in reason…
- arXiv cs.AIagentsRSI-Forge: From Research Papers to Environments for Recursive Self-Improvement
arXiv:2610.09426v1 Announce Type: new Abstract: Environments are the foundation of recursive self-improvement: they provide the problems agents work on and the feedback used to evaluate progress. Yet constructing challen…
- arXiv cs.AIresearchLLM-Enabled UAV Dispatch: A System-Level Survey and Taxonomy
arXiv:2610.09466v1 Announce Type: new Abstract: Unmanned aerial vehicle (UAV) dispatch is beginning to move beyond isolated path planning and optimization-driven resource allocation toward system-level coordination suppo…
- arXiv cs.AIagentsMIMESIS: Learning User Simulators as Training Environments for Interactive Agents
arXiv:2610.09484v2 Announce Type: replace Abstract: Training and evaluating interactive language agents typically requires rich user interactions, yet collecting human feedback is expensive and difficult to scale. Simula…
- arXiv cs.AIresearchThe Attribution Blind Spot: Layerwise Trajectory Diagnostics for Source Reliance in Retrieval-Augmented Language Models
arXiv:2610.09493v1 Announce Type: new Abstract: A retrieval-augmented model can match a document without relying on it. Controlled knowledge conflicts make source choice observable and let us ask a second question that p…
- arXiv cs.AIresearchCognitive Schemas, Laws and Tasks
arXiv:2610.09495v1 Announce Type: new Abstract: This paper asks how explicit representations can support reusable cognitive schemas in knowledge-based problem solving. We develop a structural framework in which schemas a…
- arXiv cs.AIagentsReam: Unfolding Mutual Awareness in Human-Agent Workspaces
arXiv:2610.09497v1 Announce Type: new Abstract: As AI agents work alongside humans in shared workspaces, a mutual awareness challenge arises: agents act at speeds that outpace human monitoring, and users' evolving intere…
- arXiv cs.AIresearchReliability of LLM Judges for Evaluating Entity Alignment
arXiv:2610.09554v2 Announce Type: replace Abstract: Entity Alignment (EA) identifies equivalent entities across knowledge graphs and is critical for knowledge base integration and ontology merging. Evaluating EA systems…
- arXiv cs.AIagentsDrugTargetWorld: A Synthetic Biobank for Training and Benchmarking AI Scientists
arXiv:2610.09558v1 Announce Type: new Abstract: Drug target discovery requires distinguishing molecules that causally drive disease from those that are merely associated with it. Training and evaluating AI agents to perf…
- arXiv cs.AIagentsWorld Potential Model: Pretrained World Knowledge as Progress Potentials
arXiv:2610.09560v1 Announce Type: new Abstract: Long-horizon language agents often receive supervision only from terminal task outcomes, leaving little signal for distinguishing productive intermediate behavior from stag…
- arXiv cs.AIresearchDual- versus Single-Suggestion AI Support for Radiographic Interpretation in Residents: Randomized Multireader Study
arXiv:2610.09589v1 Announce Type: new Abstract: Purpose: To compare dual- and single-suggestion AI support for radiographic interpretation by residents, particularly when the shared AI suggestion was incorrect. Materials…
- arXiv cs.AIagentsLearning Situation-Conditioned Thinking Policies for Long-Term LLM Agents
arXiv:2610.09590v1 Announce Type: new Abstract: Long-running autonomous agents must reuse accumulated reasoning experience without allowing explicit historical memory and LLM context to grow indefinitely. However, existi…
- arXiv cs.AIresearchSafeEvo: Deciphering the Safety Alignment Mechanism and Evolution in Language Models
arXiv:2610.09600v1 Announce Type: new Abstract: Safety interpretability advances the study of Large Language Model (LLM) alignment from behavioral constraints driven by data or algorithms towards a deeper understanding o…
- arXiv cs.AIagentsHow Do Agentic LLMs Decide to Call Tools? A Tool-Call Vector Shaped by Suppression
arXiv:2610.09624v2 Announce Type: replace Abstract: Tool calling, invoking external tools on demand, is central to agentic LLMs, yet the mechanism that decides whether a model calls a tool or responds directly remains po…
- arXiv cs.AIresearchAutomatically Building and Updating a Knowledge Graph of MLIP Models
arXiv:2610.09644v1 Announce Type: new Abstract: Complementing the many efforts in providing semantic representations of concepts, notions, and entities in materials science, we report and illustrate a process by which we…
- arXiv cs.AIagentsMeshSIPP: Efficient Lattice Planning in Dynamic Environment
arXiv:2610.09652v1 Announce Type: new Abstract: Autonomous navigation in dynamic environments requires computing spatiotemporal trajectories that satisfy non-holonomic motion constraints. When the trajectories of the mov…
- arXiv cs.AIagentsShared and structured inputs undermine collective random choice by reasoning AI agents
arXiv:2610.09667v1 Announce Type: new Abstract: Random selection is widely used in resource allocation and auditing, making reliable implementation essential for AI-agent systems. Behavioural tests across six reasoning m…
- arXiv cs.AIagentsCost-Efficient Theorem Proving via Agent Orchestration in Program Verification
arXiv:2610.09681v1 Announce Type: new Abstract: Program verification establishes software correctness through machine-checkable proofs constructed in theorem provers. It's a guarantee especially valuable for code generat…
- arXiv cs.AIagentsSystem Switch: When Should a Fast Decision Model Stop and Think?
arXiv:2610.09683v1 Announce Type: new Abstract: Dual-process agents pair a fast policy with a slow deliberative model. In real-time settings the slow model usually runs continuously; in turn-based agents and robot planne…
- arXiv cs.AIresearchA Tale of Two Error Categories: Exploring Concealed Trade-Offs in the Errors of Automated Judges in Evaluation of Uncertainty Quantifiers
arXiv:2610.09693v1 Announce Type: new Abstract: The wide adoption of LLMs across broad NLG applications heightens the importance of providing users with the means to avert errors and hallucinations. Uncertainty quantific…
- arXiv cs.AIresearchBolzano: From Expert-Guided Proof Search to Automated Open-Problem Solving
arXiv:2610.09769v2 Announce Type: replace Abstract: Large language models are increasingly contributing to mathematical research, where progress often depends on efficient proof search, incremental improvements and caref…
- arXiv cs.AIagentsSkillForge: Co-Evolving Skills and Agents via Dynamic Skill Lifecycles
arXiv:2610.09832v2 Announce Type: replace Abstract: Memory-augmented reinforcement learning strengthens LLM agents' ability to solve complex long-horizon tasks. Skills are one such form of memory, pairing instructions wi…
- arXiv cs.AIagentsSelf-Evolve With a Reference:Anchored Training of Tool-Integrated Agents
arXiv:2610.09856v1 Announce Type: new Abstract: Self-evolving tool-integrated agents learn from tasks and feedback generated within their own training loop. A Curriculum Agent generates tasks, while an Executor Agent lea…
- arXiv cs.AIagentsLiveMACE: Process-Aware Evaluation of LLM Agent Capabilities in Evolving Markets
arXiv:2610.09872v1 Announce Type: new Abstract: Evaluating agents by outcomes alone can obscure the capabilities that produce them. This problem is especially pronounced in evolving environments, where outcomes reflect a…
- arXiv cs.AIresearchThink Before You Paint: Recursive Latent Reasoning for Diffusion Models
arXiv:2610.09876v1 Announce Type: new Abstract: Diffusion models generate realistic images but often fail on visual reasoning tasks, such as filling in a Sudoku or drawing the path through a maze. When a discrete symboli…
- arXiv cs.AIagentsOutperformance Inverse Optimization: Learning Objective Functions that Outperform Agent Decisions
arXiv:2610.09890v1 Announce Type: new Abstract: Inverse optimization estimates the weights of an objective function that explain observed decisions as optimal solutions, and is used in a variety of fields. For mixed-inte…
- arXiv cs.AIresearchNL2Hull: A Natural Language-Driven Constrained Ship Design Decision Framework
arXiv:2610.09896v1 Announce Type: new Abstract: Ship-form design combines smooth geometric representation, local shape editing, and constraints on the resulting hull. We present the Natural-Language-to-Hull Framework (NL…
- arXiv cs.AIresearchWhere Can a Decision Model Diagnose HVAC Faults? Reasoning Demand, Physical Representation, and Robustness Under Shift
arXiv:2610.09937v1 Announce Type: new Abstract: Artificial intelligence supports building operations in several forms, each with its own barrier. Expert rules must be tuned for every system, supervised models need labele…
- arXiv cs.AIagentsAgentTime: Can Agents Estimate and Control Their Own Runtime?
arXiv:2610.09944v2 Announce Type: replace Abstract: An essential control of AI agents is their ability to manage runtime. This ability requires a sense of time-awareness, to predict and estimate wall-clock time and to co…
- arXiv cs.AIresearchSuccessive Training Stages and Large Language Model Persuasion: Effects of Misalignment, Supervised Fine-Tuning, and Preference Optimization
arXiv:2610.09964v1 Announce Type: new Abstract: Large language models (LLMs) can be tuned to influence human attitudes, yet the respective contributions of successive post-training stages remain un-clear. This study exam…
- arXiv cs.AIagentsFrom Expected Harmfulness to Likelihood: A Probabilistic Reformulation of Jailbreaking LLM Agents
arXiv:2610.09973v1 Announce Type: new Abstract: When the harmfulness of an LLM agent's output can be quantified, a natural jailbreaking objective is to maximize expected harmfulness over admissible input modifications. A…
- arXiv cs.AIresearchWhat the Sleeve Feels: Explainable Machine Learning for Textile Pressure-Based Postural Screening
arXiv:2610.10015v1 Announce Type: new Abstract: Pressure-sensing smart textiles convert body-surface contact into a dense, image-like signal closely tied to posture and movement, making them a promising low-cost route to…
- arXiv cs.AIagentsLearning to Accumulate Knowledge with Mutual Information
arXiv:2610.10042v1 Announce Type: new Abstract: Large language model (LLM) agents can improve their performance by reusing knowledge distilled from past interactions. However, curating new experiences into a knowledge ba…
- arXiv cs.AIagentsLoud Failures, Quiet Failures: Fault Detection and Recovery in Tool-Using Language Model Agents
arXiv:2610.10062v1 Announce Type: new Abstract: Tool-using agents are usually scored on whether they finish a task while the tools work. Deployments are less forgiving: services time out, endpoints disappear, parameter n…
- arXiv cs.AIagentsHGP:An on-device personalized agent memory via hybrid graph storage
arXiv:2610.10071v1 Announce Type: new Abstract: LLM-based agents face challenges in personalized interactive tasks due to heterogeneous, multi-typed, and implicitly constrained long-term traces. Existing memory mechanism…
- arXiv cs.AIagentsSkillSandbox: Skill Verification via Dynamic Scenario Synthesis
arXiv:2610.10088v1 Announce Type: new Abstract: Self-evolving agents distill task-solving experience into skills for future reuse, but these skills can encode incorrect procedures or non-transferable knowledge. It is the…
- arXiv cs.AIagentsRewardWeaver: Long-Horizon Interactive Learning for Language Agents via Self-Evolving Reward Adaptation
arXiv:2610.10120v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards (RLVR) has driven substantial progress in domains where task outcomes can be reliably evaluated, but long-horizon interaction…
- arXiv cs.AIagentsUniSkill: Learning Actor-Aligned Skill Proposals for an Evolving Policy
arXiv:2610.10164v1 Announce Type: new Abstract: Large language model agents can improve across tasks by retaining reusable skills distilled from prior interactions. Recent work jointly optimizes task execution and skill…
- arXiv cs.AIagentsAgentic AI-Assisted Modeling for Production Scheduling: Assessment in Constraint Programming
arXiv:2610.10184v1 Announce Type: new Abstract: Developing optimization models for production scheduling requires substantial expert effort. Research on large language models (LLMs) has followed two directions: specializ…
- arXiv cs.AIresearchGAGR-Lab: Evaluating Joint Spatial-Geometric and Analytic Function Reasoning
arXiv:2610.10201v1 Announce Type: new Abstract: Joint spatial-geometric and analytic function reasoning requires translating a perceived spatial configuration into a symbolic function whose executed curve satisfies geome…
- arXiv cs.AIagentsOOM-RL II: Reality Is an Oracle, Not a Debugger Provenance-Constrained Diagnosis in Continually Evolving Agent-Engineered Systems
arXiv:2610.10256v1 Announce Type: new Abstract: Reality may establish that an outcome occurred without identifying which evolving procedure produced it or why. This distinction matters in production ML systems whose code…
- arXiv cs.AIagentsStale, Misattributed, or Late: Where Personal Memory Fails Before Generation
arXiv:2610.10265v1 Announce Type: new Abstract: Personal memory for language agents is usually judged by whether the final an- swer is correct. That score hides errors that arise before generation: the memory block may c…
- arXiv cs.AIagentsAI Safety Considerations for Agents With Limited Time to Act
arXiv:2610.10285v1 Announce Type: new Abstract: In the wake of the increasingly public discussion about AI alignment, recent work has tried to propose specific AI architectures that behave safely. However, the proposed a…
- arXiv cs.AIresearchThe Handover Problem: Governing Autonomy Transitions in Human-AI Collaboration
arXiv:2610.10352v1 Announce Type: new Abstract: Human-machine systems rarely operate at a fixed level of AI autonomy. As operators and AI systems collaborate over time, control must shift: the AI can take on more respons…
- arXiv cs.AIresearchOpen-MMUnlearning: Unifying Methods and Evaluation for MLLM Unlearning
arXiv:2610.10358v2 Announce Type: replace Abstract: As multimodal large language models (MLLMs) become more capable and widely deployed, concerns about privacy and safety have become increasingly pressing. Machine unlear…
- arXiv cs.AIresearchReasoning-Token Spikes Under Prompted Untruthful Responses in Large Language Models
arXiv:2610.10405v2 Announce Type: replace Abstract: Monitoring the chain-of-thought of reasoning artificial intelligence (AI) models remains a key approach to detecting deception and other forms of misbehavior in such mo…