LeanDojo and ReProver: finding useful premises in a formal library

What changed

LeanDojo extracts source-linked premises and proof states and provides programmatic Lean interaction; ReProver uses accessible-premise retrieval to improve proof generation on explicit random and novel-premise benchmark splits.

Scientific scope

LeanDojo extracts source-linked premises and proof states and provides programmatic Lean interaction; ReProver uses accessible-premise retrieval to improve proof generation on explicit random and novel-premise benchmark splits. This assesses arXiv 2306.15626v2 and its accessible data/retrieval/proof-search framework. The reported Lean 3 results distinguish random and novel-premise splits: 51.2% and 26.3%, respectively. Pass@1 here means one search attempt with a ten-minute limit, not one language-model call. Lean 4 experiments are separate. The 98,734 existing human library theorems are a research dataset, not AI discoveries. Human library construction, framework authorship and system execution remain distinct. WRS has not independently replayed the benchmark or audited every dataset statement.

Evidence

Source-linked, scoped account of the original work. The published scope states the scientific assumptions and remaining verification limits. WRS has not independently certified the full proof, experimental result or system benchmark.

Contributors

• Original9author team (complete byline retained) · SCIENCE: Toolkit/method/benchmark design, experiments and paper; no individual effort fractions inferred.
• ReProver / LeanDojo research team (source-reported) · EXECUTION: Accessible-premise retrieval and proof-search execution in reported benchmark scope; not human authorship.
• Lean/mathlib human formalizers, libraries and generic ByT5/DPR tools · INFRASTRUCTURE: Inherited scientific content and infrastructure remain separate from this method delta.
• Junyan Xu and community reviewers (source-reported roles) · VERIFICATION: Manual inspection and feedback on formalizations; not independent WRS proof replay or original paper coauthorship.

• Yang et al.2023: Lean Dojo: Theorem Proving with Retrieval-Augmented Language Models — ar Xiv 2306.15626 v 2
• mini F 2 F PR 13: 33 New Proofs Found by Machines
• Proof Net PR 14: New Proofs Generated by Machines

Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.

Full scientific record · Browse contribution grades · Request a correction