AI news archive.
4,965 articles filed under Agents, newest first, from the outlets listed on the sources page.
- arXiv cs.AIagentsWhen Numbers Start Talking: Numerical Signalling and Strategic Behaviour Among LLMs
arXiv:2610.03033v1 Announce Type: new Abstract: Large language model (LLM)-based agents increasingly operate in multi-agent systems (MAS) characterised by strategic interaction. However, little is known about whether, an…
- arXiv cs.AIagentshacktrace: behavior-supervised detection of reward hacking during code generation
arXiv:2610.03055v1 Announce Type: new Abstract: A coding agent can earn a passing grade by fixing its code, or by deleting the test that exposes the bug. Detecting such reward hacking requires recognizing attempted short…
- arXiv cs.AIagentsMOF-VERIFY: A Failure-Aware Agentic Harness for MOF Hypothesis Verification
arXiv:2610.03056v1 Announce Type: new Abstract: Large language models are increasingly used as reasoning components in AI-driven materials Co-Scientists, yet the reliability of the resulting verification pipeline remains…
- arXiv cs.AIagentsPeer Influence across Heterogeneous AI Models
arXiv:2610.03095v1 Announce Type: new Abstract: When two AI agents disagree, who persuades whom? As multi-agent systems increasingly combine language models of different families and sizes, the answer can determine which…
- arXiv cs.AIagentsToward SLM-based agentic task-tool intent matching
arXiv:2610.03213v1 Announce Type: new Abstract: Tool-equipped AI agents use tool calls to access data and act on external systems. Horizontal growth of agentic systems increases the number of these interactions, and furt…
- arXiv cs.AIagentsLightweight, Rubric-Guided Trajectory Evaluation for Production AI Agents
arXiv:2610.03315v1 Announce Type: new Abstract: Trajectory evaluation is essential for improving the reliability of LLM-based agents, but production use makes it expensive to run repeatedly. Modern agents generate long t…
- arXiv cs.AIagentsReFract: Benchmarking Perspective Awareness in Language Model Agents with Text World Models
arXiv:2610.03356v1 Announce Type: new Abstract: Large Language Model (LLM) agents are increasingly deployed in high-stakes settings such as industrial maintenance and equipment fault troubleshooting, where workers occupy…
- arXiv cs.AIagentsKnowledge or Calculator? Decomposing the Skill Premium in Verifiable Financial Agent Workflows
arXiv:2610.03564v1 Announce Type: new Abstract: Financial AI agents must do more than retrieve facts: investment workflows require correct quantitative execution, reliable use of procedural resources, and auditable struc…
- arXiv cs.AIagentsHyperBrowseComp: A Multilingual and Multimodal Stress Test for Web-Browsing Agents
arXiv:2610.03574v1 Announce Type: new Abstract: We introduce HyperBrowseComp, a multilingual and multimodal browsing benchmark comprising 423 manually authored and human-validated questions across 13 languages, written b…
- arXiv cs.AIagentsHazardWeaver: Scientific Route Selection for Hazard Analysis Agents
arXiv:2610.03591v1 Announce Type: new Abstract: Understanding and assessing natural hazards is essential for disaster preparedness and risk reduction. Recent advances in large language models have spurred growing interes…
- arXiv cs.AIagentsNeutronGym: Physics-Graded Neutron Instrument Design for LLM Agents
arXiv:2610.03631v1 Announce Type: new Abstract: Designing a scientific instrument tests whether language-model agents can do physics rather than recall it, provided the grading cannot be argued with. We introduce Neutron…
- arXiv cs.AIagentsCredit Where It Matters: Dependency-Aware Policy Optimization for Terminal Agents
arXiv:2610.03634v1 Announce Type: new Abstract: Terminal-using agents benefit from reinforcement learning (RL) in coding, debugging, and other multi-step terminal tasks. In these tasks, later commands often depend on inf…
- arXiv cs.AIagentsClarifyCodeBench: Evaluating LLMs on Clarifying Ambiguous Requirements for Code Generation
arXiv:2607.00711v2 Announce Type: cross Abstract: Large Language Models have emerged as programming assistants. However, the efficacy of code generation is constrained by the quality of input requirements, which are freq…
- arXiv cs.AIagentsCONTRA: Discovering and Qualifying Behavior-Changing Questions for Selective Clarification in LLM Code Generation
arXiv:2610.01769v1 Announce Type: cross Abstract: Coding agents can generate code that appears correct but implements behavior the user never intended. This mismatch can arise when an agent silently resolves underspecifi…
- arXiv cs.AIagentsRxnOptBench: Benchmarking LLMs for Reaction-Condition Optimization in Organic Methodology
arXiv:2610.02242v1 Announce Type: cross Abstract: Chemical reaction-condition optimization -- choosing the catalyst, ligand, solvent, reagent, temperature, time, and atmosphere that jointly maximize yield and stereoselec…
- arXiv cs.AIagentsSimuVerity: Benchmarking Agents for Engineering-Grade Simulink Model Generation
arXiv:2610.02304v1 Announce Type: cross Abstract: Existing Simulink benchmarks mainly evaluate whether generated models compile, execute, or resemble a reference model. These criteria do not establish whether a model sat…
- arXiv cs.AIagentsDeskForge: Dense Supervision from Desktop Environments for Computer-Use Agents
arXiv:2610.02320v1 Announce Type: cross Abstract: Computer-use agents need to reliably ground action targets in complex desktop scenes, where multiple applications, overlapping windows, and visually similar controls comp…
- arXiv cs.AIagentsMIRROR: Multipath Quorum Integrity for LLM Multi-Agent Communication
arXiv:2610.02349v1 Announce Type: cross Abstract: Inter-agent communication is central to Large Language Model Multi-Agent Systems (LLM-MAS), but it introduces an underexplored vulnerability: Agent-in-the-Middle (AiTM) a…
- arXiv cs.AIagentsCoco: An Agentic Copilot for the Hardware--Software Co-Design Lifecycle
arXiv:2610.02376v1 Announce Type: cross Abstract: Co-designing ML models and the accelerators that run them is an unusual reasoning task: architects must draw confident, high-stakes conclusions about systems that do not…
- arXiv cs.AIagentsInherit-MAS: Test-Time Evolution of Multi-Agent Systems through Workflow and Execution Inheritance
arXiv:2610.02396v1 Announce Type: cross Abstract: Multi-agent systems (MAS) built from large language models coordinate specialized agents to tackle complex tasks, but effective workflows are difficult to design in advan…
- arXiv cs.AIagentsSideKernel: A Usable microVM Sandbox for AI Coding Agents on macOS
arXiv:2610.02456v1 Announce Type: cross Abstract: AI coding agents are untrusted system components, yet they require autonomy on the developer machines they run on. This contradiction is a security problem. Sandboxes pro…
- arXiv cs.AIagentsAPDMem: Agent-Controlled Progressive Disclosure for Query-Adaptive Long-Term Memory
arXiv:2610.02472v1 Announce Type: cross Abstract: Personalized LLM assistants must recover sparse evidence from long conversation histories across queries of varying complexity. We introduce APDMem (Agent-controlled Prog…
- arXiv cs.AIagentsFrom Fragments to Global Maps: Learning Vectorized Map Aggregation with Large Language Models
arXiv:2610.02513v1 Announce Type: cross Abstract: Large-scale vectorized HD maps provide structured road information that is essential for perception, localization, and planning in autonomous driving. Constructing such m…
- arXiv cs.AIagentsOpenGameEval: Benchmarking Agentic Programming and Exploration in a Stateful Game Engine
arXiv:2610.02563v1 Announce Type: cross Abstract: We present OpenGameEval, a benchmark and evaluation framework for agentic game development inside Roblox Studio. It runs language models as agents in reproducible, statef…
- arXiv cs.AIagentsPincer: Resource Authorization for Agents using a Digital Twin
arXiv:2610.02569v1 Announce Type: cross Abstract: Coding agents have become increasingly long-horizon, autonomous, reliant on general-purpose shell and maintain their own persistent memory for self-improvement. While the…
- arXiv cs.AIagentsWebUIProof: Benchmarking WebUI Code Generators with UI-Agent Execution Harness
arXiv:2610.02617v1 Announce Type: cross Abstract: Evaluating WebUI code generation at scale is difficult: outputs may compile and look plausible yet fail under user interaction, and prior benchmarks largely rely on free-…
- arXiv cs.AIagentsLEAP: Learning Efficient Action Proposals For LLM Agents
arXiv:2610.02670v1 Announce Type: cross Abstract: LLM agents are known to be slow in rollouts. An agent completes a task one step at a time. At each step, it reasons and then chooses an action to execute. The next step a…
- arXiv cs.AIagentsSelf-Supervised Scaling of Terminal Environments for Scientific Domains
arXiv:2610.02710v1 Announce Type: cross Abstract: Terminal agents are increasingly deployed beyond software engineering in science and other specialized domains. Constructing training environments requires executable ref…
- arXiv cs.AIagentsProspective Hindsight: Self-Calibrating Reinforcement Learning via Prediction-Reality Gaps
arXiv:2610.02740v1 Announce Type: cross Abstract: Reinforcement learning for long-horizon agents relies on purely retrospective training signals: credit is assigned only after observing environmental consequences, leavin…
- arXiv cs.AIagentsWhen History Fails to Become Experience: Action Calibration in Language Agents
arXiv:2610.02769v1 Announce Type: cross Abstract: Language agents should draw on prior attempts and environmental feedback to improve subsequent decisions within the same task. However, providing additional interaction h…
- arXiv cs.AIagentsAgentTrap: Stateful Feedback Deception against Autonomous Penetration Testing Agents
arXiv:2610.02869v1 Announce Type: cross Abstract: Autonomous penetration testing agents conduct multi-step attacks by continuously adapting their plans and actions to target responses. As a common defense, honeypots can…
- arXiv cs.AIagentsDiscriminating Fixture Coverage in Agent-Infrastructure Verification Suites
arXiv:2610.02928v1 Announce Type: cross Abstract: Invariant suites and runtime monitors increasingly gate agent deployment decisions, and the evidence offered for any particular suite is almost always a single observatio…
- arXiv cs.AIagentsDynamic Expert Pruning for Multi-Agent Systems
arXiv:2610.02951v1 Announce Type: cross Abstract: Mixture-of-Experts (MoE) architectures scale language models efficiently by activating only a few experts per token, but the saving is confined to computation: every expe…
- arXiv cs.AIagentsSentry: Learning to Recover from LLM Agent Failures at Test Time
arXiv:2610.02994v1 Announce Type: cross Abstract: LLM agents often fail mid-task due to invalid tool calls, repeated actions, or poorly grounded reasoning, and learning from these failures is a path to reliability. We fi…
- arXiv cs.AIagentsBeyond Predefined Sinks: Security-Aware Dependency Analysis for LLM Agents
arXiv:2610.03014v1 Announce Type: cross Abstract: Large language model (LLM)-based agents increasingly connect model-generated decisions to security-sensitive software capabilities such as command execution, filesystem a…
- arXiv cs.AIagentsOmniAct3D: Leveraging Foundation Geometry and Evidence-Grounded Reasoning for Panoramic 3D Detection
arXiv:2610.03015v1 Announce Type: cross Abstract: Accurate 3D detection is essential for mobile embodied agents, while Vision Foundation Models (VFMs) offer transferable visual and geometric priors. Yet existing VFM-base…
- arXiv cs.AIagentsWebFovea: When the Model Is Right but the Click Is Wrong -- Reliable Round Trips for Vision-Based Web Agents on Live Websites
arXiv:2610.03036v2 Announce Type: replace-cross Abstract: We present WebFovea, a vision-based web agent that placed 2nd in the WebRetriever Challenge 2026 with a final score of 57.0 out of 100. The challenge evaluates ag…
- arXiv cs.AIagentsSecuring Computer-Use Agents Against Branch Steering Attacks
arXiv:2610.03089v1 Announce Type: cross Abstract: Modern Computer Use Agents (CUAs) directly interact with graphical user interfaces and execute third-party web tools, exposing them to indirect prompt injection across ev…
- arXiv cs.AIagentsAsk, Relax, or Act? Evaluating Actionable Indeterminacy in LLM Preference Reasoning
arXiv:2610.03102v1 Announce Type: cross Abstract: An LLM agent can recognize uncertainty yet still choose the wrong next step: asking when action is already justified, or seeking clarification when the constraints must c…
- arXiv cs.AIagentsEvoRiskBench: An Evolving Benchmark for Runtime Security Risks in Workspace Agents
arXiv:2610.03153v1 Announce Type: cross Abstract: Workspace agents combine large language models with execution harnesses to perform stateful, multi-step tasks that access or modify external resources. Existing benchmark…
- arXiv cs.AIagentsD2K-Bench: Can LLM Agents Turn Expert Designs into Efficient GPU Kernels?
arXiv:2610.03226v1 Announce Type: cross Abstract: GPU kernels generated by large language model (LLM) agents can remain less efficient than expert implementations, but runtime alone does not reveal how the gap relates to…
- arXiv cs.AIagentsCordial Learning: Distributed Training with Correlated Data
arXiv:2610.03330v1 Announce Type: cross Abstract: We consider a distributed learning task with agents that have correlated data. Specifically, the label of an agent depends on the input of other agents for the same sampl…
- arXiv cs.AIagentsFollow the Winners: Conservative Policy Improvement with the Cross-Entropy Method for Critic-Free RFT
arXiv:2610.03361v1 Announce Type: cross Abstract: Critic-free reinforcement fine-tuning (RFT) for agentic large language models is often done through GRPO-style methods, which compute a group baseline over repeated rollo…
- arXiv cs.AIagentsMobiAgent: Dual-Loop Recursive Policy Self-Improvement for Long-Horizon Mobile Manipulation
arXiv:2610.03476v1 Announce Type: cross Abstract: Long-horizon mobile manipulation presents significant challenges due to compounding execution errors and capacity interference between locomotion and arm control. While r…
- arXiv cs.AIagentsThreat-Preserving Representation Sensitivity in Agent-Security Benchmarks
arXiv:2610.03585v1 Announce Type: cross Abstract: Security benchmarks for LLM-based agents often report the attack success rate (ASR) as a measure of model robustness and use these scores to compare different models and…
- arXiv cs.AIagents4DCodeBench: Benchmarking Agents on Inverse Graphics of Dynamic Scenes
arXiv:2610.03715v1 Announce Type: cross Abstract: We introduce 4DCodeBench, a benchmark for 4D inverse graphics through code generation, in which agents reconstruct dynamic scenes from video as executable graphics progra…
- arXiv cs.AIagentsLPS-Bench: Benchmarking Safety Awareness of Computer-Use Agents in Long-Horizon Planning under Benign and Adversarial Scenarios
arXiv:2602.03255v2 Announce Type: replace Abstract: Computer-use agents (CUAs) execute multi-stage tasks through tools, where an early unsafe decision can propagate to consequential actions. Evaluating only final outcome…
- arXiv cs.AIagentsOne Hypothesis Is Not Enough: Abductive Reasoning with Agentic Hypothesis Refinement over Knowledge Graphs
arXiv:2605.31370v2 Announce Type: replace Abstract: Abductive reasoning over knowledge graphs (KGs) seeks a first-order logic hypothesis whose answer set explains a given set of observed entities. Since many hypotheses c…
- arXiv cs.AIagentsCounterfactual Evidence Audits Predict LLM-Agent Susceptibility to Ranked Context
arXiv:2606.00914v2 Announce Type: replace Abstract: LLM agents increasingly decide from evidence assembled by upstream systems: retrievers choose documents, recommenders choose posts, and memory systems choose prior even…
- arXiv cs.AIagentsDissecting model behavior through agent trajectories
arXiv:2606.17454v3 Announce Type: replace Abstract: AI agent performance is not just a modeling problem, it is fundamentally a systems problem. The advanced capabilities of models are realized through agent harnesses. Th…