OPSD Compresses What RLVR Teaches: A Post-RL Compaction Stage for Reasoning Models

ArXi:2605.06188v1 Announce Type: cross On-Policy Self-Distillation (OPSD) has recently emerged as an alternative to Reinforcement Learning with Verifiable Rewards (RLVR), promising higher accuracy and shorter responses through token-level credit assignment from a self-teacher conditioned on privileged context. However, this promise does not carry over to thinking-enabled mathematical reasoning, where reported accuracy gains shrink and sometimes turn negative.