AI RESEARCH
Multi-Rollout On-Policy Distillation via Peer Successes and Failures
arXiv CS.AI
•
ArXi:2605.12652v1 Announce Type: cross Large language models are often post-trained with sparse verifier rewards, which indicate whether a sampled trajectory succeeds but provide limited guidance about where reasoning succeeds or fails. On-policy distillation (OPD) offers denser token-level supervision by