Text-to-SQL on BIRD official (test)
73.67Total AccuracyRoboPhD (Best Evolved)
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| RoboPhD (Best Evolved)Model=Opus 4.5, Prompt Strategy=Best Evolved2026.01 | 73.67 | 81.03 | 70.45 | 55.44 | |
| Opus 4.5 (Naive)Model=Opus 4.5, Prompt Strategy=Naive2026.01 | 72.16 | 81.35 | 68.65 | 48.42 |