Medical Question Answering on cMedQA2 (test)
89.65Win RateLLM-AutoDP
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| LLM-AutoDPComparison Baseline=No-Process, Evaluated Model=Gemma-2-9B, Judge Model=Baichuan-M1-14B-Instruct2026.01 | 89.65 | 1.66 | 8.69 | |
| LLM-AutoDPComparison Baseline=No-Process, Evaluated Model=Llama3.1-8B, Judge Model=Baichuan-M1-14B-Instruct2026.01 | 84.29 | 4.52 | 11.19 | |
| LLM-AutoDPComparison Baseline=No-Process, Evaluated Model=Qwen2.5-7B, Judge Model=Baichuan-M1-14B-Instruct2026.01 | 81.65 | 7.78 | 10.57 | |
| LLM-AutoDPComparison Baseline=All-Process, Evaluated Model=Gemma-2-9B, Judge Model=Baichuan-M1-14B-Instruct2026.01 | 63.27 | 15.41 | 21.32 | |
| LLM-AutoDPComparison Baseline=RS, Evaluated Model=Llama3.1-8B, Judge Model=Baichuan-M1-14B-Instruct2026.01 | 61.48 | 15.93 | 22.59 | |
| LLM-AutoDPComparison Baseline=RS, Evaluated Model=Gemma-2-9B, Judge Model=Baichuan-M1-14B-Instruct2026.01 | 61.37 | 18.89 | 19.74 | |
| LLM-AutoDPComparison Baseline=All-Process, Evaluated Model=Llama3.1-8B, Judge Model=Baichuan-M1-14B-Instruct2026.01 | 60.69 | 14.3 | 25.01 | |
| LLM-AutoDPComparison Baseline=RS, Evaluated Model=Qwen2.5-7B, Judge Model=Baichuan-M1-14B-Instruct2026.01 | 60.03 | 16.83 | 23.13 | |
| LLM-AutoDPComparison Baseline=All-Process, Evaluated Model=Qwen2.5-7B, Judge Model=Baichuan-M1-14B-Instruct2026.01 | 59.55 | 15.2 | 25.25 | |
| LLM-AutoDPComparison Baseline=SELA, Evaluated Model=Qwen2.5-7B, Judge Model=Baichuan-M1-14B-Instruct2026.01 | 45.84 | 24.37 | 29.79 | |
| LLM-AutoDPComparison Baseline=SELA, Evaluated Model=Gemma-2-9B, Judge Model=Baichuan-M1-14B-Instruct2026.01 | 44.67 | 24.88 | 30.45 | |
| LLM-AutoDPComparison Baseline=SELA, Evaluated Model=Llama3.1-8B, Judge Model=Baichuan-M1-14B-Instruct2026.01 | 43.45 | 22.97 | 33.58 |