Rare Disease Diagnosis on MIMIC-IV Rare OOD (test)
53.05Top-1 AccuracyRareDxR1-Qwen + CRR
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| RareDxR1-Qwen + CRRParams=14B, Domain=Medical, RLM=Yes2026.06 | 53.05 | 62.2 | 69.83 | |
| RareDxR1-DeepSeek + CRRParams=14B, Domain=Medical, RLM=Yes2026.06 | 52.03 | 60.34 | 65.59 | |
| RareDxR1-QwenParams=14B, Domain=Medical, RLM=Yes2026.06 | 48.98 | 57.8 | 65.25 | |
| RareDxR1-DeepSeekParams=14B, Domain=Medical, RLM=Yes2026.06 | 47.63 | 55.25 | 61.19 | |
| Baichuan-M1Params=14B, Domain=Medical, RLM=Yes2026.06 | 45.42 | 51.86 | 56.44 | |
| Doubao-1.5-Thinking-ProParams=-, Domain=General, RLM=Yes2026.06 | 44.58 | 50.34 | 54.75 | |
| DeepSeek-R1Params=671B, Domain=General, RLM=Yes2026.06 | 42.71 | 48.64 | 54.58 | |
| Qwen3Params=14B, Domain=General, RLM=Yes2026.06 | 40.34 | 45.08 | 50.17 | |
| DeepSeek-DistillParams=14B, Domain=General, RLM=Yes2026.06 | 39.83 | 45.76 | 48.31 | |
| Gemini-2.5-FlashParams=-, Domain=General, RLM=Yes2026.06 | 38.14 | 45.25 | 48.98 | |
| HuatuoGPT-o1Params=70B, Domain=Medical, RLM=Yes2026.06 | 34.24 | 37.97 | 44.07 | |
| GPT-4oParams=-, Domain=General, RLM=No2026.06 | 29.15 | 33.39 | 37.97 |