Scientific Multimodal Reasoning on Galaxy-10
57.72AccuracyGPT-5
Evaluation Results
| Method | Links | |
|---|---|---|
| GPT-52026.04 | 57.72 | |
| S1-VL-32B-RLParameter Count=32B, Training Phase=RL2026.04 | 51.52 | |
| Intern-S1Parameter Count=235B+6B2026.04 | 48.37 | |
| S1-VL-32B-SFTParameter Count=32B, Training Phase=SFT2026.04 | 42.45 | |
| Intern-S1-miniParameter Count=8B2026.04 | 39.06 | |
| Qwen3-VL-235B-A22B-ThinkingParameter Count=235B-A22B, Reasoning Mode=Thinking2026.04 | 33.93 | |
| Qwen3-VL-32B-ThinkingParameter Count=32B, Reasoning Mode=Thinking2026.04 | 30.44 | |
| Thyme-VLParameter Count=7B2026.04 | 27.84 | |
| Gemini 2.5 Pro2026.04 | 25.14 | |
| Gemini 2.5 Flash2026.04 | 15.78 |