Scientific Discovery Pairwise Preference on Shared 40-task (evaluation)
40Win CountStatefulDiscovery
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| StatefulDiscoveryComparison=vs. OpenEvolve2026.06 | 40 | 0 | 0 | |
| StatefulDiscoveryComparison=vs. AutoDiscovery2026.06 | 32 | 6 | 2 | |
| StatefulDiscoveryComparison=vs. SAGA2026.06 | 31 | 7 | 2 | |
| StatefulDiscoveryComparison=vs. Raw agent2026.06 | 25 | 6 | 9 |