Jul 2026
Agent Reinforcement Learning via Pivotal-Aware Self-Feedback Retry
PivoARL is proposed, a self-feedback retry framework for experience exploitation in LLM agents that identifies the pivotal erroneous turn through structured reflection and performs local retry only from the corresponding pivotal state, thereby reusing the correct prefix and reducing redundant interactions.
Weiyang Guo, Zesheng Shi, Longhui Zhang et al.
· arXiv.org · 2 citations