List your company

Scientific Reasoning on SciKnowEval Chemistry

94.3mean@16DRIFT
-2.908822.328147.56572.8019May 27, 2026Jun 1, 2026Jun 7, 2026Jun 12, 2026Jun 18, 2026Jun 23, 2026Jun 29, 2026

Evaluation results

MethodLinks
DRIFTModel=Qwen3-8B2026.0694.3
GRPOModel=Qwen3-8B2026.0693.5
SDPOModel=Qwen3-8B2026.0689.4
GRPOModel Size=Qwen3-8B, Training Domain=Material2026.0546.87
GRPOModel Size=Qwen3-8B, Training Domain=Physics2026.0546.12
ROSDModel Size=Qwen3-4B, Training Domain=Material2026.0545.48
ROSDModel Size=Qwen3-4B, Training Domain=ToolUse2026.0545.43
GRPOModel Size=Qwen3-4B, Training Domain=Biology2026.0545.23
SDPOModel Size=Qwen3-8B, Training Domain=ToolUse2026.0545.18
ROSDModel Size=Qwen3-8B, Training Domain=ToolUse2026.0545.03
SDPOModel Size=Qwen3-8B, Training Domain=Material2026.0544.07
SDPOModel Size=Qwen3-8B, Training Domain=Physics2026.0544
GRPOModel Size=Qwen3-4B, Training Domain=ToolUse2026.0543.97
SDPOModel Size=Qwen3-4B, Training Domain=Physics2026.0543.63
GRPOModel Size=Qwen3-4B, Training Domain=Physics2026.0543.6
ROSDModel Size=Qwen3-4B, Training Domain=Physics2026.0543.43
GRPOModel Size=Qwen3-4B, Training Domain=Material2026.0542.88
GRPOModel Size=Qwen3-8B, Training Domain=ToolUse2026.0542.47
ROSDModel Size=Qwen3-8B, Training Domain=Material2026.0542.12
ROSDModel Size=Qwen3-8B, Training Domain=Physics2026.0541.62
GRPOModel Size=Qwen3-8B, Training Domain=Biology2026.0541.58
ROSDModel Size=Qwen3-4B, Training Domain=Biology2026.0540.93
SDPOModel Size=Qwen3-4B, Training Domain=Material2026.0540.03
SDPOModel Size=Qwen3-4B, Training Domain=Biology2026.0539.35
ROSDModel Size=Qwen3-8B, Training Domain=Biology2026.0538.21
SDPOModel Size=Qwen3-8B, Training Domain=Biology2026.0534.78
SDPOModel Size=Qwen3-4B, Training Domain=ToolUse2026.050.83