Machine Unlearning on TOFU (Leakage Metrics)
0Leakage@1 (ES)RULE-GradDiff
Evaluation results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| RULE-GradDiffDecoding strategy=T = 1.0, p = 1.02025.11 | 0 | 0 | 0 | 0.1 | 0.1 | 0.2 | 0.3 | 0.4 | 39.8 | |
| PMCDecoding strategy=T = 1.0, p = 1.02025.11 | 1 | 1.7 | 2.9 | 4.7 | 7 | 9.6 | 12.3 | 14.9 | 23.5 |