List your company

Medical Reasoning on MedQA-USMLE Medical Non-MCQ

22.6AccuracyData Reformulation
16.77618.28819.821.312Oct 10, 2025

Evaluation results

MethodLinks
Data Reformulation2025.1022.6
CLARITYModel Size=7B2025.1022.6
PRM RL2025.1022.2
GPT-4o2025.1022.2
Standard RL2025.1019.4
CLARITYModel Size=3B2025.1019.1
Refine-then-Monitor2025.1018.3
Qwen2.5Model Size=3B2025.1017.4
Qwen2.5Model Size=7B2025.1017