Revisiting Reinforcement Learning with Verifiable Rewards from a Contrastive Perspective 事件

Revisiting Reinforcement Learning with Verifiable Rewards from a Contrastive Perspective arXiv:2605.12969v3 Announce Type: replace-cross Abstract: Group Relative Policy Optimization (GRPO) is one of the most widely adopted RLVR algorithms for post-training large language models on reasoning tasks. We first show that GRPO admits an equivalent discriminative reformulation, in which policy optimization maximizes the expected score gap between verified positive and negative rollouts. This reformula