List your company

Science Reasoning on GPQA (Pass@1 and Pass@16)

52Pass@1Base Model
31.30436.67742.0547.423Apr 20, 2026

Evaluation results

MethodLinks
Base ModelModel Scale=Qwen3-4B, Thinking Mode=Thinking2026.045284.4
MIXED-CUTSModel Scale=Qwen3-4B, Thinking Mode=Non-thinking2026.0450.184.5
GRPOModel Scale=Qwen3-4B, Thinking Mode=Non-thinking2026.0448.184.6
Base ModelModel Scale=Qwen3-4B, Thinking Mode=Non-thinking2026.0445.383.9
MIXED-CUTSModel Scale=Qwen3-1.7B, Thinking Mode=Non-thinking2026.043679.4
Base ModelModel Scale=Qwen3-1.7B, Thinking Mode=Thinking2026.0434.975.3
GRPOModel Scale=Qwen3-1.7B, Thinking Mode=Non-thinking2026.0434.280.1
Base ModelModel Scale=Qwen3-1.7B, Thinking Mode=Non-thinking2026.0432.180.3