Science Diagram Reasoning on AI2D
79.92AccuracyCoreset
Evaluation Results
| Method | Links | |
|---|---|---|
| CoresetBackbone=Qwen2.5-VL-7B, Training=finetuned2025.10 | 79.92 | |
| RECAPBackbone=Qwen2.5-VL-7B, Training=finetuned2025.10 | 78.21 | |
| UniformBackbone=Qwen2.5-VL-7B, Training=finetuned2025.10 | 76.43 | |
| VLAA-Thinker-7B2025.10 | 75.45 | |
| PropMixBackbone=Qwen2.5-VL-7B, Training=finetuned2025.10 | 75.32 | |
| Reasoning-onlyBackbone=Qwen2.5-VL-7B, Training=finetuned2025.10 | 74.97 | |
| LwFBackbone=Qwen2.5-VL-7B, Training=finetuned2025.10 | 73.93 | |
| Qwen2.5-VL-7Bvariant=Base model2025.10 | 67.62 | |
| OpenVLThinker-7B2025.10 | 59.94 | |
| Vision-R1-7B2025.10 | 0 |