Machine Unlearning on Tiny-ImageNet Forget 10%
61.03Test AccuracyBASE
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| BASEBackbone=Swin-T2026.03 | 61.03 | 86.4 | 19.58 | |
| NEGGRADBackbone=Swin-T2026.03 | 61.02 | 86.43 | 19.58 | |
| RETRAINBackbone=Swin-T2026.03 | 59.27 | 50.3 | 0 | |
| SSDBackbone=Swin-T2026.03 | 55.87 | 84.36 | 19.41 | |
| REGUNBackbone=Swin-T2026.03 | 52.72 | 49.86 | 3.05 | |
| AMUNBackbone=Swin-T2026.03 | 51.06 | 55.35 | 5.93 | |
| FINETUNEBackbone=Swin-T2026.03 | 51 | 62.4 | 6.32 | |
| ℓ1-SPARSEBackbone=Swin-T2026.03 | 50.26 | 61 | 6.39 | |
| SALUNBackbone=Swin-T2026.03 | 49.88 | 55.77 | 7.03 | |
| NEGGRAD+Backbone=Swin-T2026.03 | 46.49 | 57.18 | 9.07 |