| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| MUSE Books | UNDIAL | Privacy Leakage-89.6 | 83 | 19d ago | |
| TOFU Forget 10% | Aggregation Score65 | 81 | 15d ago | ||
| WMDP | CAP(Ours) | Bio Accuracy24.8 | 74 | 1mo ago | |
| RWKU | CAP | ASG4.8 | 74 | 1mo ago | |
| TOFU (5%) | OPT-OUT | Forget Quality86.6 | 59 | 7d ago | |
| MNIST | ManiF-SMC | Model Accuracy99.59 | 56 | 9d ago | |
| TOFU Forget 1% | GUARD-IT | Aggregation Score66 | 54 | 20d ago | |
| TOFU 1.0 (Forget10) | IHL | Model Utility (MU)52 | 53 | 1mo ago | |
| TOFU forget05 1.0 | Model Utility (MU)75.85 | 53 | 20d ago | ||
| TOFU 1.0 (forget01) | ASU | Average Score80.1 | 53 | 20d ago | |
| Tiny-ImageNet 20 classes | Retain-only Retrain | Retain Accuracy66.38 | 48 | 1mo ago | |
| CIFAR-100 10 classes | Retain-only Retrain | Retain Accuracy76.5 | 48 | 1mo ago | |
| CIFAR-100 1 class | Retain-only Retrain | Retain Accuracy76.01 | 48 | 1mo ago | |
| CIFAR-10 3 classes | Retain-only Retrain | Retain Accuracy95.37 | 48 | 1mo ago | |
| CIFAR-10 1 class | Retain-only Retrain | Retain Accuracy94.74 | 48 | 1mo ago | |
| Tiny-ImageNet 1 class (forget) | Retain-only Retrain | Retain Accuracy66.52 | 48 | 22d ago | |
| TOFU Forget01 (1% authors) | Original | Forget Quality (Rouge-L)0.99 | 48 | 2mo ago | |
| CIFAR-100 | NegGrad + SAM 0.1 | ToW (High)82.86 | 48 | 2mo ago | |
| ImageNette gas pump Class 7 (test) | SalUn | Forget Accuracy100 | 48 | 3mo ago | |
| CIFAR-10 bird, Class 2 (test) | Forgetting Accuracy (Class)100 | 48 | 3mo ago | ||
| TOFU 1.0 (Retain Set) | ROUGE-L100 | 48 | 14d ago | ||
| CIFAR-10 | Accf99.94 | 45 | 3mo ago | ||
| CIFAR-100 (test) | MaGA | Retain Acc92.34 | 45 | 2d ago | |
| TOFU 1.0 (Real Author) | ROUGE-L93 | 45 | 14d ago | ||
| Tiny-ImageNet (train) | Forgetting Accuracy (Train)99.9 | 43 | 2mo ago |
