Clinical order generation on MedChain (test)
37.88PrecisionMedResearcher-R1
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| MedResearcher-R1Category=Agentic Reasoning Methods2026.05 | 37.88 | 40.09 | 30.31 | |
| Tongyi DeepResearchCategory=Agentic Reasoning Methods2026.05 | 32.15 | 35.79 | 24.13 | |
| ReflecToolCategory=Single-Agent Methods, Backbone=Qwen2.5-7B-Instruct2026.05 | 28.73 | 38.96 | 29.91 | |
| MDAgentsCategory=Multi-Agent Methods, Backbone=Qwen2.5-7B-Instruct2026.05 | 27.62 | 45.18 | 30.3 | |
| CAREAgentCategory=Agentic Reasoning Methods, Backbone=Qwen2.5-7B-Instruct2026.05 | 26.78 | 51.73 | 31.75 | |
| ReActCategory=Single-Agent Methods, Backbone=Qwen2.5-7B-Instruct2026.05 | 26.69 | 23.03 | 21.84 | |
| AgentClinicCategory=Multi-Agent Methods, Backbone=Qwen2.5-7B-Instruct2026.05 | 21.13 | 21.25 | 18.97 | |
| MedAgentsCategory=Multi-Agent Methods, Backbone=Qwen2.5-7B-Instruct2026.05 | 18.85 | 36.28 | 21.82 |