Research map / Vision-language and multimodal / Multimodal LLMs
Integrating Visual Interpretation and Linguistic Reasoning for Geometric Problem Solving
Zixian Guo, Ming Liu, Qilong Wang, Zhilong Ji, Jinfeng Bai, Lei Zhang, Wangmeng Zuo
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 GPT-4o description preprocessing (ablation) | Ours SFT · MathVerse Text Dominant | accuracy: 47.9 -> 52.3 | Improved |
| add GPT-4o description preprocessing (ablation) | Ours SFT · MathVerse Vision Only | accuracy: 36.7 -> 39.7 | Improved |
| add GPT-4o description preprocessing (ablation) | Ours SFT · MathVerse Vision Dominant | accuracy: 38.8 -> 42.6 | Improved |
| use decoupled interpreter and reasoner instead of single LVLM (ablation) | Qwen2VL-7B · MathVerse Vision Dominant | accuracy: 39.1 -> 47.2 | Improved |
| use decoupled interpreter and reasoner instead of single LVLM (ablation) | Qwen2VL-7B · MathVerse Vision Only | accuracy: 37.3 -> 43.8 | Improved |
| use decoupled interpreter and reasoner instead of single LVLM (ablation) | Qwen2VL-7B · MathVerse Text Dominant | accuracy: 51.1 -> 54.3 | Improved |
| add stage 2 RL of LVLM (ablation) | Ours after SFT · MathVerse Vision Only | accuracy: 39.7 -> 42.8 | Improved |
| add stage 2 RL of LVLM (ablation) | Ours after SFT · MathVerse Text Dominant | accuracy: 52.3 -> 51.5 | Hurt |
| add stage 3 RL of LLM (ablation) | Ours after SFT · MathVerse Text Lite | accuracy: 46.9 -> 49.8 | Improved |
| add stage 3 RL of LLM on top of stage 2 (ablation) | Ours stage 1+2 · MathVerse Vision Dominant | accuracy: 45.5 -> 47.2 | Improved |
| add stage 3 RL of LLM on top of stage 2 (ablation) | Ours stage 1+2 · MathVerse Text Lite | accuracy: 49.8 -> 49.0 | Hurt |
| decoupled reasoning vs LLaVA-CoT end-to-end | LLaVA-CoT-11B · MathVerse | accuracy: 36.2 -> 42.5 | Improved |
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.
- LLaVA-CoT-11B — LLaVA-CoT: Let Vision Language Models Reason Step-by-Step (ICCV 2025) · better on 4 of 4 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
- Benchmarking Multimodal Mathematical Reasoning with Explicit Visual Dependency (NeurIPS 2026 · similarity 0.90)
- GeoTikzBridge: Advancing Multimodal Code Generation for Geometric Perception and Reasoning (CVPR 2026 · similarity 0.90)
- GeoX: Geometric Problem Solving Through Unified Formalized Vision-Language Pre-training (ICLR 2025 · similarity 0.90)
- Multi-view Relational Distillation for Spatial Reasoning with Vision-Language Models (NeurIPS 2026 · similarity 0.90)
- MMFineReason: Closing the Multimodal Reasoning Gap via Open Data-Centric Methods (NeurIPS 2026 · similarity 0.89)
- Insight-V: Exploring Long-Chain Visual Reasoning with Multimodal Large Language Models (CVPR 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.
