Semifactual Credit-Augmented Policy Optimization
Reinforcement learning with verifiable rewards (RLVR) has improved the reasoning capabilities of large language models (LLMs), yet their predictions remain sensitive to task-irrelevant prompt features. We investigate this sensitivity through semifactual prompt interventions that preserve the underlying problem and its answer. Our analysis reveals substantial variation in token-level sensitivity…
We haven't written up this one. arXiv cs.AI has the full story — the link below goes straight to it.
