Reproduce “LeanDojo-v2: A Comprehensive Library for AI-Assisted Theorem Proving in Lean” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Ryan Hsiang, William Adkisson, Robert Joseph George, Anima Anandkumar

LeanDojo-v2 brings data extraction, model training, inference, and proof generation into one library, while Lean4Code provides a graphical editor for related tools. It extracts proofs from Lean repositories to support training; models can suggest tactics that Lean checks or generate whole proofs. The authors report no benchmark or proof-rate result, so the paper establishes an integrated workflow, not improved theorem-solving.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.