Natural Language Inference on MultiMed-X TH
71.33AccuracyMED-COREASONER
Evaluation Results
| Method | Links | |
|---|---|---|
| MED-COREASONERbackbone=GPT-5.12026.01 | 71.33 | |
| GPT-5.12026.01 | 70 | |
| GPT-4o2026.01 | 70 | |
| GPT-5.22026.01 | 68 | |
| Claude-3.5-haiku2026.01 | 63.33 |
| Method | Links | |
|---|---|---|
| MED-COREASONERbackbone=GPT-5.12026.01 | 71.33 | |
| GPT-5.12026.01 | 70 | |
| GPT-4o2026.01 | 70 | |
| GPT-5.22026.01 | 68 | |
| Claude-3.5-haiku2026.01 | 63.33 |