LeanGeo: Formalizing Competitional Geometry problems in Lean
On the map: LLM reasoning → Spatial and embodied reasoning
Abstract (arXiv)
Geometry problems are a crucial testbed for AI reasoning capabilities. Most existing geometry solving systems cannot express problems within a unified framework, thus are difficult to integrate with other mathematical fields. Besides, since most geometric proofs rely on intuitive diagrams, verifying geometry problems is particularly challenging. To address these gaps, we introduce LeanGeo, a unified formal system for formalizing and solving competition-level geometry problems within the Lean 4 theorem prover. LeanGeo features a comprehensive library of high-level geometric theorems with Lean's foundational logic, enabling rigorous proof verification and seamless integration with Mathlib. We also present LeanGeo-Bench, a formal geometry benchmark in LeanGeo, comprising problems from the International Mathematical Olympiad (IMO) and other advanced sources. Our evaluation demonstrates the capabilities and limitations of state-of-the-art Large Language Models on this benchmark, highlighting the need for further advancements in automated geometric reasoning. We open source the theorem library and the benchmark of LeanGeo at https://github.com/project-numina/LeanGeo/tree/master.
Key facts. Cited by 7 papers in the map (data through 14 Oct 2026). Placed in Spatial and embodied reasoning, one of 17 directions in LLM reasoning. Builds on 16 of its 48 references that are in the map. First posted on arXiv 20 Aug 2025.
Builds on
16 of 48 references.
Closest ancestors · 2
Same direction, newest first
- 2024Proposing and solving olympiad geometry with guided tree search14 Dec
- 2022UniGeo: Unifying Geometry Logical Reasoning via Reformulating Mathematical Expression6 Dec
Other references · 14
- 2025Qwen3 Technical Reportcited by 8,820
- 2025Kimi k1.5: Scaling Reinforcement Learning with LLMscited by 1,210
- 2025DeepSeek-Prover-V2: Advancing Formal Mathematical Reasoning via Reinforcement Learning for Subgoal Decompositioncited by 287
- 2021GeoQA: A Geometric Question Answering Benchmark Towards Multimodal Numerical Reasoningcited by 228
- 2025Kimina-Prover Preview: Towards Large Formal Reasoning Models with Reinforcement Learningcited by 180
- 2025Goedel-Prover-V2: Scaling Formal Theorem Proving with Scaffolded Data Synthesis and Self-Correctioncited by 154
- 2023LEGO-Prover: Neural Theorem Proving with Growing Librariescited by 109
- 2025BFS-Prover: Scalable Best-First Tree Search for LLM-based Automatic Theorem Provingcited by 84
- 2025FormalMATH: Benchmarking Formal Mathematical Reasoning of Large Language Modelscited by 48
- 2024GeoEval: Benchmark for Evaluating LLMs and Multi-Modal Models on Geometry Problem-Solvingcited by 45
- 2025Solving Formal Math Problems by Decomposition and Iterative Reflectioncited by 35
- 2025CombiBench: Benchmarking LLM Capability for Combinatorial Mathematicscited by 31
2 more
Built on by
Citing papers per year: 2025: 1; 2026: 6.
7 papers.
- 2026DreamProver: Evolving Transferable Lemma Libraries via a Wake-Sleep Theorem-Proving Agentcited by 5
- 2026LAMP: Lean-based Agentic framework with MCP and Proof Repaircited by 1
- 2026From Solvers to Research: Large Language Model-Driven Formal Mathematics at the Research Frontiercited by 1
- 2026MIRROR: Learning from the Other View for Multi-Modal Reasoningcited by 1
- 2025An Ontology-Based Approach to Optimizing Geometry Problem Sets for Skill Developmentcited by 0
- 2026FVSpec: Real-World Property-Based Tests as Lean Challengescited by 0
- 2026Euclid-Omni : A Unified Neuro-Symbolic Framework for Plane Geometrycited by 0
- 2026MIRROR: Learning from the Other View for Multi-Modal Reasoning23 Jul 2026
- 2026From Solvers to Research: Large Language Model-Driven Formal Mathematics at the Research Frontier8 Jul 2026
- 2026LAMP: Lean-based Agentic framework with MCP and Proof Repair27 Jun 2026
- 2026Euclid-Omni : A Unified Neuro-Symbolic Framework for Plane Geometry17 Jun 2026
- 2026FVSpec: Real-World Property-Based Tests as Lean Challenges31 May 2026
- 2026DreamProver: Evolving Transferable Lemma Libraries via a Wake-Sleep Theorem-Proving Agent29 Apr 2026
- 2025An Ontology-Based Approach to Optimizing Geometry Problem Sets for Skill Development2 Sep 2025
Similar, not cited
Close in wording, no citation either way
- 2023FormalGeo: An Extensible Formalized Framework for Olympiad Geometric Problem Solving
- 2026MechGeo: Autoformalizing and Proving Euclidean Geometry in Lean 4
- 2024FGeo-TP: A Language Model-Enhanced Solver for Geometry Problems
- 2026CAM-Bench: A Benchmark for Computational and Applied Mathematics in Lean
- 2026Construction-Verification: A Benchmark for Applied Mathematics in Lean 4
- 2024Wu's Method can Boost Symbolic AI to Rival Silver Medalists and AlphaGeometry to Outperform Gold Medalists at IMO Geometry
- 2025GenesisGeo: Technical Report
- 2026Solving Is Not Drawing: A Benchmark for Diagrammatic Reasoning in Olympiad Geometry