Research map / Vision-language and multimodal / Multimodal LLMs
VisionMath: Vision-Form Mathematical Problem-Solving
Zongyang Ma, Yuxin Chen, Ziqi Zhang, Zhongang Qi, Chunfeng Yuan, Shaojie Zhu, Chengxiang Zhuo, Bing Li, Ye Liu, Zang Li, Ying Shan, Weiming Hu
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 the mathematical OCR stage (ablation) | VisionMath · Vision-MathVerse overall | accuracy (%): 37.5 -> 41.1 | Improved |
| add the mathematical OCR stage (ablation) | VisionMath · VisionMath-Bench Plane Geometry | accuracy (%): 37.2 -> 39.0 | Improved |
| add the mathematical OCR stage (ablation) | VisionMath · VisionMath-Bench Algebra | accuracy (%): 31.3 -> 34.3 | Improved |
| add figure captioning in stage 2 (ablation) | VisionMath · Vision-MathVerse overall | accuracy (%): 36.7 -> 38.2 | Improved |
| add mathematical educational video learning in stage 2 (ablation) | VisionMath · Vision-MathVerse overall | accuracy (%): 36.7 -> 39.8 | Improved |
| combine figure captioning and educational video learning (ablation) | VisionMath · Vision-MathVerse overall | accuracy (%): 39.8 -> 41.1 | Improved |
| add figure captioning in stage 2 (ablation) | VisionMath · VisionMath-Bench Plane Geometry | accuracy (%): 37.0 -> 38.0 | Improved |
| add educational video learning (ablation) | VisionMath · VisionMath-Bench Solid Geometry | accuracy (%): 32.2 -> 34.6 | Improved |
| add Chinese vision-form data to English-only tuning (ablation) | VisionMath · VisionMath-Bench Plane Geometry | accuracy (%): 24.0 -> 39.0 | Improved |
| add English vision-form data to Chinese-only tuning (ablation) | VisionMath · Vision-MathVerse overall | accuracy (%): 26.1 -> 41.1 | Improved |
| add English data to Chinese-only tuning (ablation) | VisionMath · VisionMath-Bench Plane Geometry | accuracy (%): 38.2 -> 39.0 | Mixed |
| add Chinese data to English-only tuning (ablation) | VisionMath · Vision-MathVerse overall | accuracy (%): 39.2 -> 41.1 | Improved |
Closest papers by meaning
- MV-MATH: Evaluating Multimodal Math Reasoning in Multi-Visual Contexts (CVPR 2025 · similarity 0.91)
- VisioMath: Benchmarking Figure-based Mathematical Reasoning in LMMs (ICLR 2026 · similarity 0.90)
- Benchmarking Multimodal Mathematical Reasoning with Explicit Visual Dependency (NeurIPS 2026 · similarity 0.90)
- Measuring Multimodal Mathematical Reasoning with MATH-Vision Dataset (NeurIPS 2024 · similarity 0.90)
- DynaMath: A Dynamic Visual Benchmark for Evaluating Mathematical Reasoning Robustness of Vision Language Models (ICLR 2025 · similarity 0.90)
- MINT-CoT: Enabling Interleaved Visual Tokens in Mathematical Chain-of-Thought Reasoning (NeurIPS 2025 · similarity 0.89)
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.
