None defined yet.
OmniOPD: Logit-Free On-Policy Distillation via Speculative Verification
Distilling LLM Feedback for Lean Theorem Proving