List your company

Commonsense Reasoning on GPQA (pass@1)

24.2pass@1GTPO
21.622.27522.9523.625Nov 18, 2025

Evaluation results

MethodLinks
GTPOBase Model=Qwen2.5-7B2025.1124.2
GRPOBase Model=Qwen2.5-7B2025.1121.7