Clinical Diagnosis and Reasoning on SSDF-BENCH
90ME ScoreGPT-5.2-2025-12-11
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| GPT-5.2-2025-12-11Type=General LLMs, Reasoning Capability=with2026.03 | 90 | 82 | 58.19 | 67.94 | 60.51 | 66.63 | |
| SSDF-Core-DPOType=OURS, Reasoning Capability=with2026.03 | 89 | 97 | 87.59 | 82.31 | 74.57 | 82.41 | |
| Biancang-14BType=TCM-Specific LLMs, Reasoning Capability=without2026.03 | 88 | 97 | 75.07 | 72.75 | 45.42 | 66.23 | |
| ShizhenGPT-32B-LLMType=TCM-Specific LLMs, Reasoning Capability=without2026.03 | 86 | 96 | 75.92 | 67.33 | 50.87 | 67.2 | |
| DeepSeek-V3.2Type=General LLMs, Reasoning Capability=with2026.03 | 81 | 89 | 81.43 | 83.58 | 59.94 | 73.98 | |
| GLM-4.5Type=General LLMs, Reasoning Capability=with2026.03 | 77 | 96 | 78.03 | 78.25 | 60.63 | 72.8 | |
| HuatuoGPT-o1Type=Medical LLMs, Reasoning Capability=with2026.03 | 77 | 96 | 64.17 | 67.67 | 48.23 | 62.96 | |
| Qwen3-32BType=General LLMs, Reasoning Capability=with2026.03 | 75 | 97 | 76.51 | 70.69 | 53.23 | 67.93 | |
| Sunsimiao-Qwen2-7BType=TCM-Specific LLMs, Reasoning Capability=without2026.03 | 75 | 89 | 74.97 | 58.05 | 31.02 | 55.41 | |
| Baichuan-M3Type=Medical LLMs, Reasoning Capability=with2026.03 | 74 | 77 | 77.68 | 75.08 | 59.04 | 69.27 | |
| Gemini-3Type=General LLMs, Reasoning Capability=with2026.03 | 51 | 70 | 86.17 | 81.58 | 60.1 | 69.69 | |
| MMedLM-8BType=Medical LLMs, Reasoning Capability=without2026.03 | 19 | 66 | 9.45 | 9.27 | 13.09 | 17.48 | |
| Zhongjing-LLaMA-loraType=TCM-Specific LLMs, Reasoning Capability=without2026.03 | 17 | 57 | 6.37 | 4.98 | 4.56 | 11.49 |