Interclass Confusion on CIFAR10
0.06Interclass Confusion CostSSD
Evaluation results
| Method | Links | |
|---|---|---|
| SSDNum. Corrupted Samples=500, Unlearning Mode=Corrective Unlearning with replacement2026.04 | 0.06 | |
| SSDNum. Corrupted Samples=2500, Unlearning Mode=Corrective Unlearning with replacement2026.04 | 0.06 | |
| SSDNum. Corrupted Samples=5000, Unlearning Mode=Corrective Unlearning with replacement2026.04 | 0.06 | |
| RELOADNum. Corrupted Samples=5000, Unlearning Mode=Corrective Unlearning with replacement2026.04 | 0.24 | |
| CFNum. Corrupted Samples=500, Unlearning Mode=Corrective Unlearning with replacement2026.04 | 0.25 | |
| CFNum. Corrupted Samples=2500, Unlearning Mode=Corrective Unlearning with replacement2026.04 | 0.27 | |
| CFNum. Corrupted Samples=5000, Unlearning Mode=Corrective Unlearning with replacement2026.04 | 0.27 | |
| ScrubNum. Corrupted Samples=500, Unlearning Mode=Corrective Unlearning with replacement2026.04 | 0.27 | |
| ScrubNum. Corrupted Samples=2500, Unlearning Mode=Corrective Unlearning with replacement2026.04 | 0.27 | |
| ScrubNum. Corrupted Samples=5000, Unlearning Mode=Corrective Unlearning with replacement2026.04 | 0.27 | |
| RELOADNum. Corrupted Samples=500, Unlearning Mode=Corrective Unlearning with replacement2026.04 | 0.35 | |
| RELOADNum. Corrupted Samples=2500, Unlearning Mode=Corrective Unlearning with replacement2026.04 | 0.46 | |
| BadTNum. Corrupted Samples=500, Unlearning Mode=Corrective Unlearning with replacement2026.04 | 0.62 | |
| BadTNum. Corrupted Samples=2500, Unlearning Mode=Corrective Unlearning with replacement2026.04 | 0.63 | |
| BadTNum. Corrupted Samples=5000, Unlearning Mode=Corrective Unlearning with replacement2026.04 | 0.64 | |
| RewoDNum. Corrupted Samples=500, Unlearning Mode=Corrective Unlearning with replacement2026.04 | 1 | |
| RewoDNum. Corrupted Samples=2500, Unlearning Mode=Corrective Unlearning with replacement2026.04 | 1 | |
| RewoDNum. Corrupted Samples=5000, Unlearning Mode=Corrective Unlearning with replacement2026.04 | 1 |