Medical Diagnosis on VeriSim Noisy Levels 1-3
69.2Top-1 AccuracyQwen-2.5-72B
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Qwen-2.5-72BSize=72B, Type=General2026.04 | 69.2 | -15.3 | 34.5 | |
| Llama-3.1-70BSize=70B, Type=General2026.04 | 65.5 | -16.6 | 35.3 | |
| OpenBioLLM-70BSize=70B, Type=Medical2026.04 | 63.1 | -16.1 | 36 | |
| Meditron-70BSize=70B, Type=Medical2026.04 | 62.8 | -15.6 | 35.2 | |
| BioMistral-7BSize=7B, Type=Medical2026.04 | 41.8 | -22.4 | 50 | |
| Llama-3.1-8BSize=8B, Type=General2026.04 | 40.2 | -21.6 | 48.8 | |
| Mistral-7BSize=7B, Type=General2026.04 | 33.5 | -24.5 | 55.1 |