Machine Unlearning on CIFAR-100 Random Forget (40%)
96.96FAIU
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| IUBackbone=ResNet-18, Gamma=1/42026.01 | 96.96 | 96.91 | 70.27 | 3.04 | 26.2 | |
| BEBackbone=ResNet-18, Gamma=1/42026.01 | 96.92 | 96.85 | 70.33 | 3.08 | 26.21 | |
| BSBackbone=ResNet-18, Gamma=1/42026.01 | 96.92 | 96.88 | 70.23 | 3.08 | 26.18 | |
| GABackbone=ResNet-18, Gamma=1/42026.01 | 96.88 | 96.91 | 70.44 | 3.12 | 26.2 | |
| l1-sparseBackbone=ResNet-18, Gamma=1/42026.01 | 96.26 | 96.23 | 67.75 | 3.74 | 25.39 | |
| FTBackbone=ResNet-18, Gamma=1/42026.01 | 96.24 | 99.96 | 69.46 | 3.76 | 24.88 | |
| FalWBackbone=ResNet-18, Gamma=1/42026.01 | 53.32 | 99.98 | 57.71 | 45.68 | 0.74 | |
| RetrainBackbone=ResNet-18, Gamma=1/42026.01 | 53.01 | 99.95 | 56.41 | 46.99 | 0 | |
| SFRonBackbone=ResNet-18, Gamma=1/42026.01 | 51.33 | 99.98 | 58.59 | 48.67 | 1.39 | |
| SalUnBackbone=ResNet-18, Gamma=1/42026.01 | 51.27 | 99.97 | 57.89 | 49.73 | 1.49 | |
| RLBackbone=ResNet-18, Gamma=1/42026.01 | 33.86 | 95.17 | 54.51 | 66.14 | 11.25 |