Visual Mathematical Reasoning on MathVista (Avg@3)
0.735Avg@3Vision-R1-7B
Evaluation results
| Method | Links | |
|---|---|---|
| Vision-R1-7B2026.02 | 0.735 | |
| ADORABackbone=Qwen2.5-VL-7B, RL Algorithm=GRPO + ADORA2026.02 | 0.735 | |
| MM-EUREKA-7B2026.02 | 0.727 | |
| MMR1-math-v02026.02 | 0.702 | |
| GRPOBackbone=Qwen2.5-VL-7B, RL Algorithm=GRPO2026.02 | 0.702 | |
| Qwen2.5-VL-7BRL Algorithm=None2026.02 | 0.673 | |
| Claude 3.7-Sonnet2026.02 | 0.668 | |
| ADORABackbone=Internvl3-2b, RL Algorithm=GRPO + ADORA2026.02 | 0.648 | |
| Vision-R1-LlamaV-CI-11BParameters=11B2025.03 | 0.627 | |
| Mulberry-Llama-11BParameters=11B2025.03 | 0.611 | |
| GRPOBackbone=Internvl3-2b, RL Algorithm=GRPO2026.02 | 0.607 | |
| Gemini2-flash2026.02 | 0.591 | |
| Internvl3-2bRL Algorithm=None2026.02 | 0.57 | |
| LLaVA-Cot-11BParameters=11B2025.03 | 0.548 | |
| Llama-3.2-11B-VParameters=11B2025.03 | 0.486 | |
| ADORABackbone=Gemma3-4b-it, RL Algorithm=GRPO + ADORA2026.02 | 0.483 | |
| GRPOBackbone=Gemma3-4b-it, RL Algorithm=GRPO2026.02 | 0.472 | |
| Gemma3-4b-itRL Algorithm=None2026.02 | 0.463 |