Natural Language Inference on MultiMed-X SW
73.33AccuracyMED-COREASONER
Evaluation Results
| Method | Links | |
|---|---|---|
| MED-COREASONERbackbone=GPT-5.12026.01 | 73.33 | |
| GPT-5.12026.01 | 72.67 | |
| GPT-5.22026.01 | 70 | |
| GPT-4o2026.01 | 69.33 | |
| Claude-3.5-haiku2026.01 | 56.67 |
| Method | Links | |
|---|---|---|
| MED-COREASONERbackbone=GPT-5.12026.01 | 73.33 | |
| GPT-5.12026.01 | 72.67 | |
| GPT-5.22026.01 | 70 | |
| GPT-4o2026.01 | 69.33 | |
| Claude-3.5-haiku2026.01 | 56.67 |