Medical Reasoning on MedBrowseComp
35.1Accuracy (Before)Direct
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Direct2026.05 | 35.1 | 35.1 | 0 | |
| Self-Refine2026.05 | 34.9 | 33.5 | 1.4 | |
| Self-Reflection2026.05 | 34.9 | 31.4 | 3.5 | |
| CausalFlow2026.05 | 30.8 | 61.6 | 30.8 |
| Method | Links | |||
|---|---|---|---|---|
| Direct2026.05 | 35.1 | 35.1 | 0 | |
| Self-Refine2026.05 | 34.9 | 33.5 | 1.4 | |
| Self-Reflection2026.05 | 34.9 | 31.4 | 3.5 | |
| CausalFlow2026.05 | 30.8 | 61.6 | 30.8 |