DeepSeek-Prover-V1.5 Harnessing Proof Assistant Feedback for Reinforcement Learning and Monte-Carlo Tree Search discuss: https://
huggingface.co/papers/2408.08
152
… We introduce DeepSeek-Prover-V1.5, an open-source language model designed for theorem proving in Lean 4, which enhances
DeepSeek-Prover-V1.5: Advanced Theorem Proving with Reinforcement Learning
By
–
