Machine Unlearning on CIFAR-20 Lamp
85.52RABT
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| BTBackbone=ResNet-18, RTE=292024.02 | 85.52 | 10 | 84.84 | 0.72 | 0 | |
| PretrainBackbone=ResNet-18, RTE=69102024.02 | 85.31 | 74.22 | 85.21 | 21.3 | 92.82 | |
| RetrainBackbone=ResNet-18, RTE=42982024.02 | 85.12 | 11.31 | 84.4 | 0 | 7.06 | |
| SSDBackbone=ResNet-18, RTE=182024.02 | 84.56 | 15 | 83.84 | 1.6 | 0.6 | |
| SalUnBackbone=ResNet-18, RTE=10072024.02 | 84.44 | 13.7 | 83.74 | 1.24 | 1.68 | |
| MU-MisBackbone=ResNet-18, RTE=102024.02 | 84.36 | 11.7 | 83.66 | 0.63 | 0 | |
| DUCKBackbone=ResNet-18, RTE=682024.02 | 83.25 | 31.02 | 82.69 | 7.75 | 27.68 | |
| FTBackbone=ResNet-18, RTE=1282024.02 | 82.47 | 14 | 81.9 | 1.97 | 2.8 | |
| SCRUBBackbone=ResNet-18, RTE=1132024.02 | 82.17 | 19 | 81.6 | 4.48 | 26.2 | |
| MunBaBackbone=ResNet-18, RTE=6762024.02 | 81.17 | 17 | 80.69 | 4.45 | 5 | |
| LoTusBackbone=ResNet-18, RTE=142024.02 | 27.44 | 12 | 27.4 | 38.45 | 33.6 |