Multimodal Reasoning on MathVerse (Pass@1)
67.84Pass@1Self-distill Masking-KD-8B
Evaluation Results
| Method | Links | |
|---|---|---|
| Self-distill Masking-KD-8BModel Size=~8B, Decoding Strategy=greedy, Max Length=4096, Distillation=Self-distill2026.05 | 67.84 | |
| Qwen3-VL-8B-ThinkingModel Size=~8B, Decoding Strategy=greedy, Max Length=40962026.05 | 63.81 | |
| Masking-KD-4BModel Size=~4B, Decoding Strategy=greedy, Max Length=4096, Distillation=Distilled from 8B2026.05 | 62.66 | |
| Ovis2-8BModel Size=~8B, Decoding Strategy=greedy, Max Length=40962026.05 | 61.19 | |
| Ovis2-4BModel Size=~4B, Decoding Strategy=greedy, Max Length=40962026.05 | 58.03 | |
| Masking-KD-2BModel Size=~2B, Decoding Strategy=greedy, Max Length=4096, Distillation=Distilled from 8B2026.05 | 57.89 | |
| InternVL3-8BModel Size=~8B, Decoding Strategy=greedy, Max Length=40962026.05 | 54.82 | |
| InternVL3.5-8BModel Size=~8B, Decoding Strategy=greedy, Max Length=40962026.05 | 53.26 | |
| Ovis2-2BModel Size=~2B, Decoding Strategy=greedy, Max Length=40962026.05 | 50.32 | |
| Qwen3-VL-4B-ThinkingModel Size=~4B, Decoding Strategy=greedy, Max Length=40962026.05 | 49.86 | |
| InternVL3.5-4BModel Size=~4B, Decoding Strategy=greedy, Max Length=40962026.05 | 43.21 | |
| Qwen2.5-VL-3BModel Size=~4B, Decoding Strategy=greedy, Max Length=40962026.05 | 40.69 | |
| InternVL3-2BModel Size=~2B, Decoding Strategy=greedy, Max Length=40962026.05 | 40.37 | |
| Qwen2.5-VL-7BModel Size=~8B, Decoding Strategy=greedy, Max Length=40962026.05 | 38.67 | |
| Qwen3-VL-2B-ThinkingModel Size=~2B, Decoding Strategy=greedy, Max Length=40962026.05 | 28.21 | |
| InternVL3.5-2BModel Size=~2B, Decoding Strategy=greedy, Max Length=40962026.05 | 27.94 | |
| MiMo-VL-8BModel Size=~8B, Decoding Strategy=greedy, Max Length=40962026.05 | 18.12 |