Shypd.aiThe largest open index of AI toolsDeepSeek-Prover-V1.5 is an open-source language model for theorem proving in Lean 4. It uses proof assistant feedback and Monte-Carlo Tree Search to enhance formal verification.
Trending
About
DeepSeek-Prover-V1.5 is an open-source language model specifically designed for theorem proving within the Lean 4 proof assistant environment. It significantly enhances its predecessor, DeepSeek-Prover-V1, by optimizing both training and inference processes. The model is pre-trained on DeepSeekMath-Base and specializes in formal mathematical languages. It undergoes supervised fine-tuning with an advanced formal theorem proving dataset and further refinement through reinforcement learning from proof assistant feedback (RLPAF). A key innovation is RMaxTS, a Monte-Carlo tree search variant that employs an intrinsic-reward-driven exploration strategy to generate diverse proof paths. This approach has led to state-of-the-art results on benchmarks like miniF2F and ProofNet.
Capabilities
Pricing & Plans
Open Source
Free
FAQs
Trending
