Self-Retrospection Distillation: Turning Post-hoc Experiences into Prior Foresight

Excerpt

arXiv:2610.08077v1 Announce Type: cross Abstract: Reinforcement learning with verifiable rewards (RLVR) turns agent experience into learning signals primarily through scalar outcome rewards after interaction. For group-relative objectives, however, this signal vanishes when all rollouts receive the same reward, even though their trajectories may reveal useful information about what the task requires and how the agent fails. We ask a complementary question: can hindsight teach an agent what it co