tool · first seen 4 Oct 2026
Lean theorem prover
Also: Lean, lean4, Lean 4, Lean formal proof, Lean formalization, Lean 4 formalization, Lean proof assistant
Interactive theorem prover and programming language used for formal verification of code and mathematics.
Lean is an interactive theorem prover and functional programming language in which proofs are checked by a small trusted kernel; Lean 4 and its Mathlib library are the current standard. The sources show it serving as the verification target for LLM-driven work: agentic autoformalization of research mathematics, certified planning heuristics, proof refinement for analytic constants, and a statistics prover-library. Large formalizations (finite simple group classification, the Stacks Project) from Axiom Math, a Terence Tao essay on reliability, an interview with creator Leo de Moura, and applications to system identification and the US tax code round out the activity.
Lean has been widely used for years, so the spike reflects its growing role as the checker that makes AI-generated mathematics and code verifiable, not a new idea. Worth tracking is the gap the sources flag between a proof that compiles and a formal statement that faithfully matches the original claim.
Analyst note (established), 17 Nov 2026.
Mentions per week
PapersCodeDiscussion
Rank 128 on 14 Nov 2026 · 45 mentions in 28 days · 44 independent groups · papers → code → discussion · first seen 6 weeks ago · acceleration -0.77
Earliest sources
- AIProver: Agentic Auto-Formalization of Mathematical Research via Certificate-Driven Evolving HarnessPapers · 4 Oct 2026 · Prithwish Jana, Viet Bach Hoang, Logan Luna, Viresh Pati, Akash Singirikonda, Cy…
- scottnarmstrong/LeanAutoformalizationSkillsCode · 4 Oct 2026 · scottnarmstrong · 72 stars
- LeanPlan: Optimal Planning with LLM-Generated Heuristics and Admissibility ProofsPapers · 6 Oct 2026 · André G. Pereira, Augusto B. Corrêa, Felipe Meneguzzi, Jendrik Seipp
- SCOPE: Certified Theorem Proving with a Language Model as the Policy PlannerPapers · 6 Oct 2026 · Hanchao Zhou, Jialei Li
- An AI-Assisted Formalization of the Poincaré ConjecturePapers · 6 Oct 2026 · Zhiyuan Zhang, Axel Delaval, Leheng Chen, Jinxuan Chen, Jie Xu, Yuxuan Liao, Jie…
- SpecGuard: Proving a Task Is Broken Before the Agent CheatsPapers · 6 Oct 2026 · Param Biyani, Krishnamurthy Dvijotham
Latest sources
- Interpreting claims that a theorem is correct because it was checked in LeanDiscussion · 16 Nov 2026 · matt_d · 317 points
- leochlon/tiny-bayesCode · 16 Nov 2026 · leochlon · 22 stars
- Formalizing US Tax Code (Using Lean)Discussion · 14 Nov 2026 · dnw · 152 points
- From Pen-and-Paper to Machine-Checked Proofs: Lean for System IdentificationPapers · 13 Nov 2026 · Manas Mejari, Dario Piga, Marco Forgione
- Trust, but verify, audit and attribute: a human-in-the-loop framework for checking AI-produced research resultsPapers · 9 Nov 2026 · György Eigner, Patrik P. Süli, Roland Hollós
- Statopia: A Prover--Library Ecosystem for Automated Formalization and Proof Discovery in StatisticsPapers · 8 Nov 2026 · Ziyuan Chen, Tong Gan, Wei Zhang, Ruofei Fang, Boyue Tang, Zhenhua Lin, Fang Yao
- Improving Analytic Constants via Interpretable AI-Guided Lean Proof RefinementPapers · 7 Nov 2026 · Rahul Saha, Alan Li, Anton Xue, Swarat Chaudhuri, Adam Klivans, Pravesh K. Kotha…
- The Stacks Project, Formalized in LeanDiscussion · 5 Nov 2026 · namanbnsl · 97 points
- AxiomMath/StacksCode · 4 Nov 2026 · AxiomMath · 139 stars
- Formalizing inverse problems in the age of AI: possibilities, pitfalls, and an invitationPapers · 4 Nov 2026 · Christian Clason, Tapio Helin, Medet Nursultanov, Lauri Oksanen
- dhsorens/fv-workshopCode · 3 Nov 2026 · dhsorens · 20 stars
- Whiel: A Neurosymbolic Platform for Certified Relational Program VerificationPapers · 3 Nov 2026 · Jesse Comer, Fangzhu Shen, Shuo Zhang, Mayur Naik, Sudeepa Roy, Val Tannen