List your company

Scientific Reasoning on SciKnowEval Physics

63.39Mean@16GRPO
-2.02614.95731.9448.923May 27, 2026

Evaluation results

MethodLinks
GRPOModel Size=Qwen3-8B, Training Domain=Material2026.0563.39
GRPOModel Size=Qwen3-8B, Training Domain=Chemistry2026.0561.8
GRPOModel Size=Qwen3-8B, Training Domain=Biology2026.0560.89
GRPOModel Size=Qwen3-4B, Training Domain=Material2026.0560.78
GRPOModel Size=Qwen3-8B, Training Domain=ToolUse2026.0557.76
GRPOModel Size=Qwen3-4B, Training Domain=ToolUse2026.0557.29
ROSDModel Size=Qwen3-8B, Training Domain=Biology2026.0556.98
GRPOModel Size=Qwen3-4B, Training Domain=Biology2026.0556.35
ROSDModel Size=Qwen3-8B, Training Domain=Material2026.0555.57
ROSDModel Size=Qwen3-8B, Training Domain=ToolUse2026.0554.3
GRPOModel Size=Qwen3-4B, Training Domain=Chemistry2026.0549.71
ROSDModel Size=Qwen3-4B, Training Domain=Biology2026.0549.58
ROSDModel Size=Qwen3-8B, Training Domain=Chemistry2026.0549.09
ROSDModel Size=Qwen3-4B, Training Domain=ToolUse2026.0546.25
SDPOModel Size=Qwen3-4B, Training Domain=Biology2026.0545.68
ROSDModel Size=Qwen3-4B, Training Domain=Material2026.0545.05
SDPOModel Size=Qwen3-8B, Training Domain=Material2026.0542.76
SDPOModel Size=Qwen3-8B, Training Domain=Biology2026.0540.36
SDPOModel Size=Qwen3-8B, Training Domain=ToolUse2026.0540.34
ROSDModel Size=Qwen3-4B, Training Domain=Chemistry2026.0539.95
SDPOModel Size=Qwen3-4B, Training Domain=Material2026.0530.29
SDPOModel Size=Qwen3-8B, Training Domain=Chemistry2026.0529.92
SDPOModel Size=Qwen3-4B, Training Domain=Chemistry2026.0523.05
SDPOModel Size=Qwen3-4B, Training Domain=ToolUse2026.050.49