Repairs for Whom? Evaluator-Executor Mismatch in Selecting Repairs for Failed LLM Agents
cs.SE, cs.CL
Submitted: 2026-10-28
Updated: 2026-11-12
Comments: 17 pages, 4 figures, 11 tables
Code: https://github.com/lyj20071013/repairs-f
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- CausalFlow: Causal Attribution and Counterfactual Repair for LLM Agent Failures
- Process Reward Models for LLM Agents: Practical Framework and Directions
- The Replay Gap: Static Evaluation of Model Switching in LLM Agents Scores the Wrong World
- The Llama 3 Herd of Models
- Counterfactual Rollout Replay: Forkable Environments as Free Process Rewards for Software Engineering Agents
- Who&When Pro: Can LLMs Really Attribute Failures in AI Agents?
- DoVer: Intervention-Driven Auto Debugging for LLM Multi-Agent Systems
- Can Failed Trajectories Reveal Where Agents Go Wrong? Agentic Failure Attribution Without Step-Level Labels
- Process Reward Modeling with Reliability- and Conservatism-Aware Bayesian Belief
- On the Design of Qwen3.8-Next Architecture: Evaluation, Efficiency, and Training Stability
- Qwen2.5 Technical Report
- EASE: Behavior-Adaptive Skill Curation for Self-Evolving Agents
- Calibration Is Not Control: Intervention Value for LLM-Agent Oversight
- Agent Error Dataset: Scaling 50,000 Error--Diagnosis Pairs for Failure Analysis and Error-Aware Post-Training
Related papers
- ArkEval: Benchmarking and Evaluating Automated CodeRepair for ArkTS
- TestPrism: Rethinking Test Evaluation Beyond a Single Reference
- Mining Workflow Graphs for Black-Box Boundary Testing of Conversational LLM Agents
- Falsification-Based Verification of LLM-Generated Optimization Models: Sound Test Batteries and Their Detection Limits
- GitSkills: A Dataset of Agent Skills on GitHub
- For Coding Agents, the Repository Matters