AI RESEARCH
Improving LLM Code Reasoning via Semantic Equivalence Self-Play with Formal Verification
arXiv CS.LG
•
ArXi:2604.17010v1 Announce Type: cross