Research map / Large language models / Agents, code and math
Pi-GPS: Enhancing Geometry Problem Solving by Unleashing the Power of Diagrammatic Information
Junbo Zhao, Ting Zhang, Jiayu Sun, Mi Tian, Hua Huang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add text disambiguation module (ablation) | Pi-GPS (with LLM predictor) · Geometry3K | Choice accuracy: 72.3 -> 77.8 | Improved |
| add text disambiguation module (ablation) | Pi-GPS (with LLM predictor) · Geometry3K | Completion accuracy: 63.2 -> 70.6 | Improved |
| add text disambiguation module (ablation) | Pi-GPS (traversal predictor) · Geometry3K | Choice accuracy: 70.6 -> 76.6 | Improved |
| add LLM theorem predictor (ablation) | Pi-GPS · Geometry3K | Choice accuracy: 76.6 -> 77.8 | Improved |
| add LLM theorem predictor (ablation) | Pi-GPS w/o disambiguation · Geometry3K | Choice accuracy: 70.6 -> 72.3 | Improved |
| add rectifier with general prompt (ablation) | Pi-GPS · Geometry3K | Choice accuracy: 72.3 -> 71.9 | Hurt |
| add rectifier with specific prompt (ablation) | Pi-GPS · Geometry3K | Choice accuracy: 72.3 -> 73.3 | Improved |
| add verifier to rectifier (ablation) | Pi-GPS · Geometry3K | Choice accuracy: 73.3 -> 77.8 | Improved |
| add verifier to rectifier (ablation) | Pi-GPS · Geometry3K | Completion accuracy: 64.2 -> 70.6 | Improved |
| add text disambiguation to direct solving (ablation) | o3-mini · Geometry3K | Choice accuracy: 70.4 -> 75.5 | Improved |
| use o3-mini for theorem prediction instead of direct solving (ablation) | o3-mini · Geometry3K | Choice accuracy: 75.5 -> 77.8 | Improved |
| swap theorem predictor LLM (Claude, GPT-4o, Gemini 2, o3-mini) (ablation) | Pi-GPS · Geometry3K | Choice accuracy: 77.2 to 77.8 | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- E-GPS — E-GPS: Explainable Geometry Problem Solving via Top-Down Solver and Bottom-Up Generator (CVPR 2024) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- E-GPS: Explainable Geometry Problem Solving via Top-Down Solver and Bottom-Up Generator (CVPR 2024 · similarity 0.89)
- GeoX: Geometric Problem Solving Through Unified Formalized Vision-Language Pre-training (ICLR 2025 · similarity 0.87)
- Geoint-R1: Formalizing Multimodal Geometric Reasoning with Dynamic Auxiliary Constructions (CVPR 2026 · similarity 0.87)
- GeoBench: Rethinking Multimodal Geometric Problem-Solving via Hierarchical Evaluation (ICLR 2026 · similarity 0.87)
- GNS: Solving Plane Geometry Problems by Neural-Symbolic Reasoning with Multi-Modal LLMs (AAAI 2025 · similarity 0.87)
- Hilbert-Geo: Solving Solid Geometric Problems by Neural-Symbolic Reasoning (CVPR 2026 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
