self-feedback

Tag

Cards List
#self-feedback

Agent Reinforcement Learning via Pivotal-Aware Self-Feedback Retry

arXiv cs.AI · 2026-07-07 Cached

PivoARL is a self-feedback retry framework that identifies pivotal erroneous turns in LLM agent trajectories, enabling local retry to reduce interaction costs and improve learning efficiency. It achieves significant gains over baseline methods on multiple agent and QA benchmarks.

0 favorites 0 likes
← Back to home

Submit Feedback