Mathematical Reasoning on MathVista (ALL, ARI, LOG, STA, VQA Scores)
64.2ALL ScoreAStar
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| AStarBackbone=Qwen2.5-7B, Training-free=true2025.02 | 64.2 | 63.8 | 59.5 | 69.1 | 60.9 | |
| R1-VL-7B2025.02 | 63.5 | 54.7 | 28.5 | 61.2 | 60.9 | |
| AStarBackbone=Qwen2-VL-7B, Training-free=true2025.02 | 61.7 | 60.6 | 28.6 | 68.4 | 59.6 | |
| Human2025.02 | 60.3 | 59.2 | 40.7 | 63.9 | 55.9 | |
| GPT-4o2025.02 | 60.1 | 58.4 | 27 | 69 | 49.2 | |
| URSA-8B2025.02 | 59.8 | 53.5 | 21.6 | 57.1 | 40.2 | |
| Qwen2-VL-7B2025.02 | 58.9 | 57.5 | 24.3 | 43.1 | 58.1 | |
| InternVL2-8B2025.02 | 58.3 | 56.4 | 10.8 | 68.8 | 49.7 | |
| MultiMath-7B2025.02 | 50 | 42.2 | 23.3 | 64.9 | 49.2 | |
| Math-PUMA-7B2025.02 | 47.9 | 46.2 | 21.6 | 55.8 | 30.2 | |
| Math-LLaVA-13B2025.02 | 46.6 | 40.7 | 23.3 | 42.3 | 33.5 | |
| LLaVA-NeXT-34B2025.02 | 46.5 | — | — | — | — | |
| SPHINX-8x7B2025.02 | 42.6 | 43 | 14.4 | 50.8 | 43.3 | |
| SPHINX-V2-13B2025.02 | 36.7 | 33.4 | 24.3 | 51.5 | 43 | |
| LLaVA-1.5-13B2025.02 | 27.7 | 28.6 | 10.8 | 22.9 | 30.2 | |
| MiniGPT4-7B2025.02 | 23.1 | 32 | 10.8 | 17.9 | 30.2 | |
| Random2025.02 | 17.9 | 13.8 | 13.4 | 14.3 | 26.3 |