Patient Simulation on Patient Simulation Profile Modeling Dataset
0.94RealismDEPROFILE
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| DEPROFILEBackbone Model=GPT-4o-mini2026.03 | 0.94 | 0.05 | 4.39 | |
| DEPROFILEBackbone Model=DeepSeek-V3.2-Exp2026.03 | 0.939 | 0.063 | 4.85 | |
| DEPROFILEBackbone Model=Llama-3.1-8B-Instruct2026.03 | 0.936 | 0.085 | 4.24 | |
| DEPROFILEBackbone Model=Average2026.03 | 0.932 | 0.059 | 4.58 | |
| PATIENT-ΨBackbone Model=GPT-4o-mini2026.03 | 0.927 | 0.026 | 2.95 | |
| EEYOREBackbone Model=GPT-4o-mini2026.03 | 0.925 | 0.029 | 3.08 | |
| PATIENT-ΨBackbone Model=DeepSeek-V3.2-Exp2026.03 | 0.923 | 0.046 | 4.76 | |
| EEYOREBackbone Model=DeepSeek-V3.2-Exp2026.03 | 0.922 | 0.044 | 3.77 | |
| DEPROFILEBackbone Model=Qwen-3-4B-Instruct2026.03 | 0.913 | 0.039 | 4.84 | |
| EEYOREBackbone Model=Average2026.03 | 0.91 | 0.044 | 3.92 | |
| PATIENT-ΨBackbone Model=Average2026.03 | 0.908 | 0.045 | 4.05 | |
| EEYOREBackbone Model=Qwen-3-4B-Instruct2026.03 | 0.898 | 0.02 | 4.83 | |
| EEYOREBackbone Model=Llama-3.1-8B-Instruct2026.03 | 0.896 | 0.085 | 3.99 | |
| PATIENT-ΨBackbone Model=Qwen-3-4B-Instruct2026.03 | 0.893 | 0.024 | 4.76 | |
| PATIENT-ΨBackbone Model=Llama-3.1-8B-Instruct2026.03 | 0.888 | 0.085 | 3.74 |