{"total":171,"rows":[{"id":"2603.28052","title":"Meta-Harness: End-to-End Optimization of Model Harnesses","created":"2026-03-30","cited_by_count":309,"area":null},{"id":"2603.25723","title":"Natural-Language Agent Harnesses","created":"2026-03-26","cited_by_count":61,"area":null},{"id":"2605.03546","title":"ProgramBench: Can Language Models Rebuild Programs From Scratch?","created":"2026-05-05","cited_by_count":59,"area":null},{"id":"2602.21193","title":"On Data Engineering for Scaling LLM Terminal Capabilities","created":"2026-02-24","cited_by_count":43,"area":null},{"id":"2603.04370","title":"$\\tau$-Knowledge: Evaluating Conversational Agents over Unstructured Knowledge","created":"2026-03-04","cited_by_count":39,"area":null},{"id":"2603.27538","title":"LongCat-Next: Lexicalizing Modalities as Discrete Tokens","created":"2026-03-29","cited_by_count":36,"area":null},{"id":"2606.05405","title":"Agents' Last Exam","created":"2026-06-03","cited_by_count":31,"area":null},{"id":"2607.08964","title":"Long-Horizon-Terminal-Bench: Testing the Limits of Agents on Long-Horizon Terminal Tasks with Dense Reward-Based Grading","created":"2026-07-09","cited_by_count":31,"area":null},{"id":"2604.17308","title":"SkillFlow:Benchmarking Lifelong Skill Discovery and Evolution for Autonomous Agents","created":"2026-04-19","cited_by_count":25,"area":null},{"id":"2606.07682","title":"SWE-Marathon: Can Agents Autonomously Complete Ultra-Long-Horizon Software Work?","created":"2026-06-05","cited_by_count":23,"area":null},{"id":"2604.25727","title":"Toward Scalable Terminal Task Synthesis via Skill Graphs","created":"2026-04-28","cited_by_count":22,"area":null},{"id":"2606.11926","title":"Toward Generalist Autonomous Research via Hypothesis-Tree Refinement","created":"2026-06-10","cited_by_count":22,"area":null},{"id":"2606.05922","title":"Evolving Agents in the Dark: Retrospective Harness Optimization via Self-Preference","created":"2026-06-04","cited_by_count":19,"area":null},{"id":"2608.05466","title":"Recursive Synthesis for Long-Horizon Terminal Tasks","created":"2026-08-05","cited_by_count":19,"area":null},{"id":"2609.14858","title":"Dream-RSI: Recursive Self-Improvement through Evolving Worlds","created":"2026-09-14","cited_by_count":18,"area":null},{"id":"2604.11641","title":"CodeTracer: Towards Traceable Agent States","created":"2026-04-13","cited_by_count":16,"area":null},{"id":"2604.17596","title":"Terminal Wrench: A Dataset of 331 Reward-Hackable Environments and 3,632 Exploit Trajectories","created":"2026-04-19","cited_by_count":15,"area":null},{"id":"2605.12925","title":"AgentLens: Revealing The Lucky Pass Problem in SWE-Agent Evaluation","created":"2026-05-13","cited_by_count":15,"area":null},{"id":"2605.12673","title":"Do Androids Dream of Breaking the Game? Systematically Auditing AI Agent Benchmarks with BenchJack","created":"2026-05-12","cited_by_count":14,"area":null},{"id":"2606.05080","title":"AutoLab: Can Frontier Models Solve Long-Horizon Auto Research and Engineering Tasks?","created":"2026-06-03","cited_by_count":13,"area":null},{"id":"2606.12736","title":"Benchmarking AI Agents for Addressing Scientific Challenges Across Scales","created":"2026-06-10","cited_by_count":12,"area":null},{"id":"2608.27454","title":"WikiSkill: Compiling Agent Experience into Persistent Knowledge for Skill Evolution","created":"2026-08-27","cited_by_count":12,"area":null},{"id":"2607.08716","title":"Remember When It Matters: Proactive Memory Agent for Long-Horizon Agents","created":"2026-07-09","cited_by_count":11,"area":null},{"id":"2604.07355","title":"Prediction Arena: Benchmarking AI Models on Real-World Prediction Markets","created":"2026-03-28","cited_by_count":10,"area":null},{"id":"2605.07073","title":"TeamBench: Evaluating Agent Coordination under Enforced Role Separation","created":"2026-05-08","cited_by_count":10,"area":null},{"id":"2605.22535","title":"TerminalWorld: Benchmarking Agents on Real-World Terminal Tasks","created":"2026-05-21","cited_by_count":10,"area":null},{"id":"2605.29548","title":"Why Larger Models Learn More: Effects of Capacity, Interference, and Rare-Task Retention","created":"2026-05-28","cited_by_count":10,"area":null},{"id":"2603.04601","title":"Vibe Code Bench: Evaluating AI Models on End-to-End Web Application Development","created":"2026-03-04","cited_by_count":9,"area":null},{"id":"2606.03461","title":"What Makes Interaction Trajectories Effective for Training Terminal Agents?","created":"2026-06-02","cited_by_count":9,"area":null},{"id":"2606.04455","title":"The Meta-Agent Challenge: Are Current Agents Capable of Autonomous Agent Development?","created":"2026-06-03","cited_by_count":9,"area":null},{"id":"2608.04003","title":"PAST-Bench: Benchmarking the Foundations of Recursive Self-Improvement in Personal Agents","created":"2026-08-04","cited_by_count":9,"area":null},{"id":"2608.18580","title":"FACET: Preserving Source Intent and Executable State in Terminal Task Synthesis","created":"2026-08-19","cited_by_count":9,"area":null},{"id":"2603.13428","title":"SWE-Milestone: Evaluating AI Agents on Continuous Software Evolution","created":"2026-03-13","cited_by_count":8,"area":null},{"id":"2605.09636","title":"PDEAgent-Bench: A Multi-Metric, Multi-Library Benchmark for PDE Solver Generation","created":"2026-05-10","cited_by_count":8,"area":null},{"id":"2605.16679","title":"CHI-Bench: Can AI Agents Automate End-to-End, Long-Horizon, Policy-Rich Healthcare Workflows?","created":"2026-05-15","cited_by_count":8,"area":null},{"id":"2605.29559","title":"LiteCoder-Terminal: Scaling Long-Horizon Terminal Environments for Learning Language Agents","created":"2026-05-28","cited_by_count":8,"area":null},{"id":"2608.06352","title":"CalibForge: Adversarial Solver Calibration for Scaling Learnable Terminal Tasks","created":"2026-08-06","cited_by_count":8,"area":null},{"id":"2609.04148","title":"Terminal-Universe: Turning Agent Trajectories into Scalable Terminal Environments","created":"2026-09-03","cited_by_count":8,"area":null},{"id":"2606.14502","title":"From Chatbot to Digital Colleague: The Paradigm Shift Toward Persistent Autonomous AI","created":"2026-06-12","cited_by_count":6,"area":null},{"id":"2606.30182","title":"MirrorCode: AI can rebuild entire programs from behavior alone","created":"2026-06-29","cited_by_count":6,"area":null},{"id":"2607.14004","title":"Do Agent Optimizers Compound? A Continual-Learning Evaluation on Terminal-Bench 2.0","created":"2026-07-15","cited_by_count":6,"area":null},{"id":"2603.02540","title":"A Neuropsychologically Grounded Evaluation of LLM Cognitive Abilities","created":"2026-03-03","cited_by_count":5,"area":null},{"id":"2604.06742","title":"Evaluating LLM-Based 0-to-1 Software Generation in End-to-End CLI Tool Scenarios","created":"2026-04-08","cited_by_count":5,"area":null},{"id":"2605.24117","title":"SkillEvolBench: Benchmarking the Evolution from Episodic Experience to Procedural Skills","created":"2026-05-22","cited_by_count":5,"area":null},{"id":"2606.04261","title":"Can Generalist Agents Automate Data Curation?","created":"2026-06-02","cited_by_count":5,"area":null},{"id":"2608.20169","title":"Task-CoEvolve: Efficient Harness Optimization via Adaptive Validation Task Selection","created":"2026-08-20","cited_by_count":5,"area":null},{"id":"2604.15151","title":"QuantCode-Bench: A Benchmark for Evaluating the Ability of Large Language Models to Generate Executable Algorithmic Trading Strategies","created":"2026-04-16","cited_by_count":4,"area":null},{"id":"2605.20520","title":"Open-World Evaluations for Measuring Frontier AI Capabilities","created":"2026-05-19","cited_by_count":4,"area":null},{"id":"2606.01286","title":"BenchEvolver: Frontier Task Synthesis via Solution-Centric Evolution","created":"2026-05-31","cited_by_count":4,"area":null},{"id":"2607.16387","title":"Fantastic Adaptive Taxonomies and How to Use Them","created":"2026-07-17","cited_by_count":4,"area":null},{"id":"2608.17271","title":"ASI-Bench: At the Dawn of Artificial Superintelligence","created":"2026-08-18","cited_by_count":4,"area":null},{"id":"2604.01375","title":"The Hitchhikers Guide to Rubric Quality Understanding and Enrichment","created":"2026-04-01","cited_by_count":3,"area":null},{"id":"2604.06111","title":"AgentCE-Bench: Agent Configurable Evaluation with Scalable Horizons and Controllable Difficulty under Lightweight Environments","created":"2026-04-07","cited_by_count":3,"area":null},{"id":"2604.27865","title":"KellyBench: A Benchmark for Long-Horizon Sequential Decision Making","created":"2026-04-30","cited_by_count":3,"area":null},{"id":"2605.10516","title":"Consistency as a Testable Property: Statistical Methods to Evaluate AI Agent Reliability","created":"2026-05-11","cited_by_count":3,"area":null},{"id":"2605.13880","title":"PREPING: Zero-Task Agent Memory Construction via Self-Practice","created":"2026-05-11","cited_by_count":3,"area":null},{"id":"2605.14133","title":"ClawForge: Generating Executable Interactive Benchmarks for Command-Line Agents","created":"2026-05-13","cited_by_count":3,"area":null},{"id":"2605.17172","title":"OpenJarvis: Personal AI, On Personal Devices","created":"2026-05-16","cited_by_count":3,"area":null},{"id":"2605.17829","title":"Interactive Evaluation Requires a Design Science","created":"2026-05-18","cited_by_count":3,"area":null},{"id":"2606.00341","title":"ROGUE: Evaluating Corrigibility Failures in Frontier Computer-Use Agents","created":"2026-05-29","cited_by_count":3,"area":null},{"id":"2606.03889","title":"RealClawBench: Live OpenClaw Benchmarks from Real Developer-Agent Sessions","created":"2026-06-02","cited_by_count":3,"area":null},{"id":"2606.14516","title":"Every Eval Ever: A Unifying Schema and Community Repository for AI Evaluation Results","created":"2026-06-12","cited_by_count":3,"area":null},{"id":"2607.17598","title":"Is Progressive Disclosure All You Need for Long-Context Agents?","created":"2026-07-20","cited_by_count":3,"area":null},{"id":"2609.17930","title":"Locating Hidden Failures Makes Long-Horizon Agents More Reliable","created":"2026-09-15","cited_by_count":3,"area":null},{"id":"2605.11209","title":"Measuring Five-Nines Reliability: Sample-Efficient LLM Evaluation in Saturated Benchmarks","created":"2026-05-11","cited_by_count":2,"area":null},{"id":"2605.12233","title":"No More, No Less: Task Alignment in Terminal Agents","created":"2026-05-12","cited_by_count":2,"area":null},{"id":"2605.21516","title":"Harnesses for Inference-Time Alignment over Execution Trajectories","created":"2026-05-15","cited_by_count":2,"area":null},{"id":"2605.29682","title":"Scaling Laws for Agent Harnesses via Effective Feedback Compute","created":"2026-05-28","cited_by_count":2,"area":null},{"id":"2606.17454","title":"Dissecting model behavior through agent trajectories","created":"2026-06-16","cited_by_count":2,"area":null},{"id":"2606.20510","title":"Efficient and Sound Probabilistic Verification for AI Agents","created":"2026-06-18","cited_by_count":2,"area":null},{"id":"2607.02931","title":"VERITAS: Towards a General-Purpose Replication Tool for Scientific Research","created":"2026-07-03","cited_by_count":2,"area":null},{"id":"2607.03726","title":"SelfMem: Self-Optimizing Memory for AI Agents","created":"2026-07-04","cited_by_count":2,"area":null},{"id":"2607.09739","title":"Coresets Before Score Sets: Evaluation-Unsupervised Prompt Subset Selection for LLM Benchmarks","created":"2026-07-02","cited_by_count":2,"area":null},{"id":"2607.25090","title":"Matryoshka Agent: Unfolding Sub-Agents for Long-Horizon Machine Learning Engineering","created":"2026-07-27","cited_by_count":2,"area":null},{"id":"2608.01662","title":"LongCat Sparse Attention: Taming the Lightning via Streaming-aware Hierarchical Cross-Layer Indexing","created":"2026-08-03","cited_by_count":2,"area":null},{"id":"2608.03764","title":"GDPevo: Evaluating Agent Self-Evolution on Real Business Tasks","created":"2026-08-04","cited_by_count":2,"area":null},{"id":"2608.13566","title":"Don't Claim Benchmark-Oriented Optimization Improves General Coding Capability -- Diverse Evaluation Is Required","created":"2026-06-26","cited_by_count":2,"area":null},{"id":"2608.15242","title":"LongRCA Bench: Root-Cause Localization in Long-Horizon Agent Trajectories","created":"2026-08-15","cited_by_count":2,"area":null},{"id":"2608.16798","title":"ClawGym II: Exploring Black-Box RL on Agent Harness","created":"2026-08-17","cited_by_count":2,"area":null},{"id":"2608.29381","title":"Safe to Resume? Breaking Execution Continuity of Agent Execution via Rollback","created":"2026-08-29","cited_by_count":2,"area":null},{"id":"2608.29685","title":"Last Step Matters: Early Uncertainty Cannot Predict Failure in Long-Horizon Agents","created":"2026-08-30","cited_by_count":2,"area":null},{"id":"2609.06396","title":"MetaRSI / RSI2: A Meta-Recursive Self-Improving System for Recursive Self-Improving Systems Themselves","created":"2026-09-06","cited_by_count":2,"area":null},{"id":"2604.09836","title":"COMPOSITE-Stem","created":"2026-04-10","cited_by_count":1,"area":null},{"id":"2605.10966","title":"MMTB: Evaluating Terminal Agents on Multimedia-File Tasks","created":"2026-05-08","cited_by_count":1,"area":null},{"id":"2605.21482","title":"DeepWeb-Bench: A Deep Research Benchmark Demanding Massive Cross-Source Evidence and Long-Horizon Derivation","created":"2026-05-20","cited_by_count":1,"area":null},{"id":"2606.03918","title":"Hedge-Bench: Benchmarking Agents on Hard, Realistic Tasks Pertaining to Financial Reasoning","created":"2026-06-02","cited_by_count":1,"area":null},{"id":"2606.21005","title":"Building Agent Harnesses for Scientific Curation from Multimodal Sources","created":"2026-06-19","cited_by_count":1,"area":null},{"id":"2606.22906","title":"DeepDiscovery: A Location-Inference Framework for Task-Level Repository Understanding","created":"2026-06-22","cited_by_count":1,"area":null},{"id":"2607.18239","title":"SysAdmin: Measuring Instrumental Power-Seeking in Frontier AI","created":"2026-04-10","cited_by_count":1,"area":null},{"id":"2607.19865","title":"DocOps: A Verifiable Benchmark for Autonomous Agents in Complex Document Operations","created":"2026-07-22","cited_by_count":1,"area":null},{"id":"2607.28887","title":"To Add Is Machine, To Delete Is Human: Measuring and Mitigating Deletion Avoidance in LLM Code Editing","created":"2026-07-30","cited_by_count":1,"area":null},{"id":"2608.06714","title":"The Optimizer Is the Agent: Reasoning-Driven Search across Prompts, Programs, and ML Workflows","created":"2026-08-07","cited_by_count":1,"area":null},{"id":"2608.11234","title":"InfraBench: Evaluating Infrastructure Agents Across Layers, Lifecycle, and Risk","created":"2026-07-31","cited_by_count":1,"area":null},{"id":"2608.13547","title":"QuoteBench: How Matched Scores Can Hide Command-Path Failures","created":"2026-08-13","cited_by_count":1,"area":null},{"id":"2608.15089","title":"StateM: Reaching 95.3% Raw Accuracy, or a \\$15 Frontier Run, on Terminal-Bench 2.1 via Harness Scaling","created":"2026-08-15","cited_by_count":1,"area":null},{"id":"2608.24804","title":"StarHarness: Evolving Harnesses with Stratified Search for Enterprise Environments","created":"2026-08-25","cited_by_count":1,"area":null},{"id":"2609.30325","title":"ScopeBench: Do Agents Preserve Engagement Boundaries Under Goal Pressure?","created":"2026-09-23","cited_by_count":1,"area":null},{"id":"2609.30967","title":"MoMHa: Multi-Objective Optimization of LLM Harnesses over Accuracy, Safety, and Tokens","created":"2026-09-25","cited_by_count":1,"area":null},{"id":"2609.37725","title":"Context Language Models","created":"2026-09-29","cited_by_count":1,"area":null},{"id":"2609.38345","title":"OpenCollab: A Multi-Agent Coding Framework with Programmable Collaboration and Controllable Runtime","created":"2026-09-29","cited_by_count":1,"area":null}],"next_cursor":"100"}