Model-Based Optimization on Superconductor
1.43Expected Top-1% ScoreCliqueformer
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| CliqueformerNumber of generated designs=1000, Number of top designs for averaging=10, Number of runs=5, Score normalization method=min-max2024.10 | 1.43 | — | — | — | |
| DDOMNumber of generated designs=1000, Number of top designs for averaging=10, Number of runs=5, Score normalization method=min-max2024.10 | 1.22 | — | — | — | |
| Grad.Asc.Number of generated designs=1000, Number of top designs for averaging=10, Number of runs=5, Score normalization method=min-max2024.10 | 1.13 | — | — | — | |
| RWRNumber of generated designs=1000, Number of top designs for averaging=10, Number of runs=5, Score normalization method=min-max2024.10 | 1.03 | — | — | — | |
| IOMNumber of generated designs=1000, Number of top designs for averaging=10, Number of runs=5, Score normalization method=min-max2024.10 | 1.03 | — | — | — | |
| COMsNumber of generated designs=1000, Number of top designs for averaging=10, Number of runs=5, Score normalization method=min-max2024.10 | 0.97 | — | — | — | |
| TransformerNumber of generated designs=1000, Number of top designs for averaging=10, Number of runs=5, Score normalization method=min-max2024.10 | 0.96 | — | — | — | |
| MatchOptNumber of generated designs=1000, Number of top designs for averaging=10, Number of runs=5, Score normalization method=min-max2024.10 | 0.84 | — | — | — | |
| Cliqueformernumber of designs=128, averaging=5 runs2024.10 | — | — | — | 1.42 | |
| COMsnumber of designs=128, averaging=5 runs2024.10 | — | — | — | 0.92 | |
| DDOMnumber of designs=128, averaging=5 runs2024.10 | — | — | — | 1.2 | |
| Grad.Asc.number of designs=128, averaging=5 runs2024.10 | — | — | — | 1.19 | |
| RWRnumber of designs=128, averaging=5 runs2024.10 | — | — | — | 1 | |
| Transformernumber of designs=128, averaging=5 runs2024.10 | — | — | — | 0.91 |