List your company

Medical Reasoning on Lancet (OOD)

67AccuracyPRA (Ours)
59.92861.76463.665.436Apr 10, 2026

Evaluation results

MethodLinks
PRA (Ours)Policy=Qwen3-4B-Instruct2026.0467
CoT + SCPolicy=Qwen3-4B-Instruct2026.0463.5
CoTPolicy=Qwen3-4B-Instruct2026.0462.4
RAGPolicy=Qwen3-4B-Instruct2026.0462.1
RAG + SCPolicy=Qwen3-4B-Instruct2026.0461
DirectPolicy=Qwen3-4B-Instruct2026.0460.4
Direct + SCPolicy=Qwen3-4B-Instruct2026.0460.2