Decoupling Exploration from Optimization in RLVR
Modern language models undergo reinforcement learning with verifiable rewards (RLVR) on top of already-trained checkpoints. A key promise of RLVR is the discovery of new reasoning strategies. In principle, a model can sample novel ideas absent from its prior training data. In practice, however, augmenting RLVR with str...
Saif Punjwani, Micah Goldblum
· 0 citations