List your company

Reasoning on GPQA Diamond Expert-level OOD

37.3pass@1LUFFY
24.19627.5983134.402Oct 22, 2025

Evaluation results

MethodLinks
LUFFYBackbone=Qwen2.5-Math-7B2025.1037.3
Scaf-GRPOBackbone=Qwen2.5-Math-7B2025.1037.3
Scaf-GRPOBackbone=Qwen2.5-7B-Base2025.1035.8
LUFFYBackbone=Qwen2.5-7B-Base2025.1034.4
SimpleRL-ZeroBackbone=Qwen2.5-Math-7B2025.1033.3
Oat-ZeroBackbone=Qwen2.5-Math-7B2025.1033.3
Vanilla GRPOBackbone=Qwen2.5-7B-Base2025.1033.3
Vanilla GRPOBackbone=Qwen2.5-Math-7B2025.1032.3
Base ModelBackbone=Qwen2.5-Math-7B2025.1024.7