Virtual Standardized Patient simulation on SPBench
97.17QC ScoreEasyMED
Evaluation results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| EasyMEDCategory=Prompt Strategy + Agent Framework2025.11 | 97.17 | 97.23 | 98.18 | 97.11 | 95.03 | 97.48 | 97.98 | 95.73 | 96.98 | |
| Gemini 2.5 Pro + CoTCategory=Prompt Strategy + Agent Framework, Prompt Strategy=Chain-of-Thought2025.11 | 96.61 | 94.98 | 96.63 | 95.33 | 93.99 | 95.59 | 97.61 | 95.61 | 95.77 | |
| Human SPCategory=Prompt Strategy + Agent Framework, Reference Type=Gold standard2025.11 | 95.71 | 96.47 | 98.53 | 97.85 | 98.18 | 95.1 | 98.56 | 98.22 | 97.33 | |
| Gemini 2.5 ProCategory=LLMs2025.11 | 94.72 | 95.05 | 96.04 | 95.69 | 94.39 | 93.21 | 96.04 | 95.27 | 95.04 | |
| EvoPatientCategory=Prompt Strategy + Agent Framework2025.11 | 91.87 | 96.29 | 94.59 | 95.77 | 92.67 | 90.49 | 92.32 | 92.74 | 93.33 | |
| DeepSeek-R1Category=LLMs2025.11 | 91.06 | 93.08 | 93.42 | 97.1 | 96.1 | 85.38 | 93.69 | 95.05 | 93.11 | |
| Qwen3-32BCategory=LLMs2025.11 | 88.37 | 88.37 | 89.31 | 89.31 | 88.99 | 85.89 | 89.91 | 89.61 | 88.87 | |
| Qwen3-8BCategory=LLMs2025.11 | 77.76 | 78.64 | 84.26 | 85.74 | 83.37 | 72.14 | 85.74 | 81.31 | 81.12 | |
| GPT-4oCategory=LLMs2025.11 | 75.87 | 89.24 | 88.89 | 98.11 | 91.31 | 91.66 | 90.62 | 90.62 | 89.54 |