Fetching the paper…

Formal Theorem Proving by Rewarding LLMs to Decompose Proofs Hierarchically · Around