Clinical Medical Reasoning Evaluation on MMLU ProX Health (random sample)
50Clarity ScoreMED-COREASONER
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| MED-COREASONERComparison Baseline=GPT-5.1, Measurement=Win Rate (%)2026.01 | 50 | 43.3 | 46.7 | 60 | |
| MED-COREASONERComparison Baseline=GPT-5.1, Measurement=Tie Rate (%)2026.01 | 26.7 | 30 | 43.3 | 23.3 |