Scientific Reasoning on SciKnowEval Biology
37.08Mean@16ROSD
Evaluation Results
| Method | Links | |
|---|---|---|
| ROSDModel Size=Qwen3-8B, Training Domain=Chemistry2026.05 | 37.08 | |
| ROSDModel Size=Qwen3-8B, Training Domain=ToolUse2026.05 | 37.04 | |
| GRPOModel Size=Qwen3-8B, Training Domain=Chemistry2026.05 | 36.17 | |
| GRPOModel Size=Qwen3-8B, Training Domain=ToolUse2026.05 | 36.13 | |
| SDPOModel Size=Qwen3-8B, Training Domain=Material2026.05 | 35.83 | |
| SDPOModel Size=Qwen3-8B, Training Domain=ToolUse2026.05 | 35.21 | |
| SDPOModel Size=Qwen3-8B, Training Domain=Physics2026.05 | 34.46 | |
| ROSDModel Size=Qwen3-8B, Training Domain=Physics2026.05 | 34.17 | |
| ROSDModel Size=Qwen3-4B, Training Domain=Physics2026.05 | 33.88 | |
| GRPOModel Size=Qwen3-4B, Training Domain=Material2026.05 | 33.63 | |
| SDPOModel Size=Qwen3-4B, Training Domain=Physics2026.05 | 32.96 | |
| ROSDModel Size=Qwen3-8B, Training Domain=Material2026.05 | 31.75 | |
| SDPOModel Size=Qwen3-4B, Training Domain=Chemistry2026.05 | 31.54 | |
| ROSDModel Size=Qwen3-4B, Training Domain=Material2026.05 | 31.29 | |
| GRPOModel Size=Qwen3-4B, Training Domain=ToolUse2026.05 | 30.54 | |
| ROSDModel Size=Qwen3-4B, Training Domain=Chemistry2026.05 | 30.5 | |
| ROSDModel Size=Qwen3-4B, Training Domain=ToolUse2026.05 | 29.63 | |
| GRPOModel Size=Qwen3-4B, Training Domain=Chemistry2026.05 | 28.88 | |
| GRPOModel Size=Qwen3-4B, Training Domain=Physics2026.05 | 28.71 | |
| SDPOModel Size=Qwen3-8B, Training Domain=Chemistry2026.05 | 28.67 | |
| SDPOModel Size=Qwen3-4B, Training Domain=Material2026.05 | 28.46 | |
| GRPOModel Size=Qwen3-8B, Training Domain=Material2026.05 | 27.79 | |
| GRPOModel Size=Qwen3-8B, Training Domain=Physics2026.05 | 25.96 | |
| SDPOModel Size=Qwen3-4B, Training Domain=ToolUse2026.05 | 0.88 |