Patient Simulation on Ch-PatientSim
3.939Persona ConsistencyQwen2.5-72B+MSPRP
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Qwen2.5-72B+MSPRPModel=Qwen2.5-72B, Protocol=MSPRP2026.01 | 3.939 | 3.956 | 3.97 | 3.969 | |
| DeepSeek-V3Model=DeepSeek-V32026.01 | 3.936 | 3.883 | 3.962 | 3.916 | |
| Qwen2.5-7B+MSPRPModel=Qwen2.5-7B, Protocol=MSPRP2026.01 | 3.905 | 3.918 | 3.942 | 3.937 | |
| Qwen2.5-72BModel=Qwen2.5-72B2026.01 | 3.87 | 3.896 | 3.914 | 3.91 | |
| GPT-4ominiModel=GPT-4omini2026.01 | 3.858 | 3.855 | 3.906 | 3.871 | |
| GLM-4-9BModel=GLM-4-9B2026.01 | 3.77 | 3.79 | 3.857 | 3.796 | |
| Qwen2.5-7BModel=Qwen2.5-7B2026.01 | 3.748 | 3.795 | 3.824 | 3.78 | |
| Internlm3-8BModel=Internlm3-8B2026.01 | 3.589 | 3.642 | 3.602 | 3.572 | |
| Llama-3.1-8BModel=Llama-3.1-8B2026.01 | 3.416 | 3.515 | 3.523 | 3.384 |