AgentHER: Failed Agent Trajectories Become Training Data via Hindsight Relabeling — +11.7% Gains, 2x Data Efficiency (arXiv 2603.21357)

arXiv·medium signal

AgentHER converts failed LLM agent trajectories into valid training data by recognizing that a trajectory failing goal A is often a correct demonstration for alternative goal B. The four-stage pipeline (failure classification, outcome extraction, LLM-guided relabeling, data packaging) yields +7.1-11.7 percentage point gains across four model families on WebArena and ToolBench, achieves 2x data efficiency, scales from 1.5B to 72B parameters with 97.7% relabeling precision. Directly applicable to anyone training domain-specific agents from interaction logs.