LogP optimization with similarity preservation on MOSES
1NovDiGress
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| DiGress2026.05 | 1 | 1 | 0.13 | 0.13 | |
| PURE2026.05 | 1 | 0.45 | 0.58 | 0.26 | |
| VAE reconstruction2026.05 | 0.99 | 0.19 | 0.59 | 0.11 | |
| Mol-CycleGAN2026.05 | 0.99 | 0.51 | 0.15 | 0.08 | |
| PhAMELalign=None2026.05 | 0.99 | 0.81 | 0.35 | 0.28 | |
| PhAMEwc=6, wa=32026.05 | 0.99 | 0.78 | 0.39 | 0.3 | |
| Random target class2026.05 | 0 | 1 | 0.14 | 0 | |
| Nearest target class2026.05 | 0 | 1 | 0.6 | 0 |