Medical Ethics Evaluation on MedES 1.0 (test)
60.2EK Accuracydeepseek-r1-671b
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| deepseek-r1-671bType=DeepSeek2026.01 | 60.2 | 88.4 | 89.1 | |
| deepseek-r1-7b-ragType=Ours, Method=Retrieval-Augmented Generation2026.01 | 58.6 | 82.4 | 91.2 | |
| deepseek-v3-671bType=DeepSeek2026.01 | 56.5 | 85.9 | 88.6 | |
| jingyiqianxunType=MedicalLLM2026.01 | 54.6 | 87.7 | 91.8 | |
| ernie4Type=General-purpose2026.01 | 54.5 | 78.7 | 84.3 | |
| qwen2.5-72bType=General-purpose2026.01 | 54.1 | 84.1 | 89.2 | |
| doubaoType=General-purpose2026.01 | 48.2 | 85.4 | 89.5 | |
| qwen2.5-7bType=General-purpose2026.01 | 45.2 | 73 | 82 | |
| gpt4-turboType=GPT2026.01 | 43.4 | 69.2 | 70.8 | |
| deepseek-r1-7b-sft-round5Type=Ours, SFT Round=52026.01 | 43.1 | 52.8 | 61.6 | |
| gpt4Type=GPT2026.01 | 42.8 | 65.9 | 71.1 | |
| deepseek-r1-7b-sft-round4Type=Ours, SFT Round=42026.01 | 42.2 | 52 | 59.8 | |
| deepseek-r1-7b-sft-round3Type=Ours, SFT Round=32026.01 | 41.2 | 51 | 57.4 | |
| deepseek-r1-7b-sft-round2Type=Ours, SFT Round=22026.01 | 38.9 | 48.9 | 50.5 | |
| deepseek-r1-7b-sft-round1Type=Ours, SFT Round=12026.01 | 36.7 | 45.5 | 43.7 | |
| gpt3.5Type=GPT2026.01 | 29.3 | 45.8 | 53.2 | |
| deepseek-r1-7bType=DeepSeek2026.01 | 26.8 | 34.7 | 41 | |
| huatuogpt-o1-7bType=MedicalLLM2026.01 | 3.6 | 20.4 | 15 |