List your company

Medical Reasoning on MedQA ID (in-distribution)

80.8AccuracyPRA (Ours)
60.5265.78571.0576.315Apr 10, 2026

Evaluation results

MethodLinks
PRA (Ours)Policy=Qwen3-4B-Instruct2026.0480.8
RAG + SCPolicy=Qwen3-4B-Instruct2026.0476.7
CoT + SCPolicy=Qwen3-4B-Instruct2026.0474.8
CoTPolicy=Qwen3-4B-Instruct2026.0472.7
RAGPolicy=Qwen3-4B-Instruct2026.0472.2
DirectPolicy=Qwen3-4B-Instruct2026.0461.6
Direct + SCPolicy=Qwen3-4B-Instruct2026.0461.3