| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Image Classification | Tiny-ImageNet 1.0 (test) | Accuracy (Natural)67.64 | 75 | |
| Adversarial Attack | Tiny-ImageNet (val) | Median L2 Distance0.113 | 64 | |
| Task-incremental learning | Tiny-ImageNet 20 tasks | Average Accuracy67.95 | 54 | |
| Classification | Tiny ImageNet 200 (test) | Test Accuracy66.16 | 53 | |
| Machine Unlearning | Tiny-ImageNet 1 class (forget) | Retain Accuracy66.52 | 48 | |
| Image Classification | Tiny-ImageNet-C 30% corrupted (test) | Accuracy43.7 | 45 | |
| Image Classification | Seq-Tiny-ImageNet | Final Average Accuracy69.77 | 44 | |
| Continual Learning | Tiny-ImageNet (test) | Relative Forgetting0.2724 | 36 | |
| Image Classification | Tiny-ImageNet Dirichlet-0.05 (test) | Accuracy54 | 32 | |
| Class Incremental Learning | Tiny-ImageNet 10 tasks | Accuracy49.4 | 31 | |
| Image Classification | Tiny ImageNet 64x64 (test) | Accuracy29.6 | 27 | |
| Image Classification | Tiny-ImageNet heterogeneous (test) | Accuracy54.25 | 24 | |
| Continual Learning | S-Tiny-ImageNet (last) | TIL Last Top-1 Acc82.04 | 24 | |
| Image Classification | S-Tiny-ImageNet Task-IL | Accuracy82.04 | 24 | |
| Image Classification | S-Tiny-ImageNet Class-IL | Accuracy59.99 | 23 | |
| Near-OOD Detection | Tiny ImageNet Near-OOD (test) | FPR@9526.53 | 22 | |
| Class-incremental Continual Learning | Tiny-ImageNet Split (test) | Average Accuracy (AT)59.99 | 22 | |
| Task-incremental learning | Split Tiny-ImageNet (test) | Average Accuracy (A_T)82.04 | 22 | |
| Machine Unlearning | Tiny-ImageNet 20 classes (forget) | Accuracy66.08 | 21 | |
| Machine Unlearning | Tiny-ImageNet 1 class forget (Retain set) | Accuracy66.52 | 21 | |
| Image Classification | Tiny-ImageNet LT (test) | Accuracy (Dogs)94.67 | 20 | |
| Machine Unlearning | Tiny-Imagenet Random Forget 50%, γ=0 (test) | MIA4.32 | 19 | |
| Class-Incremental Learning | Tiny-ImageNet Split | Accuracy25.15 | 17 | |
| Image Classification Unlearning | Tiny ImageNet 30% class-wise forgetting | Accuracy (Train Retained)92.7 | 16 | |
| Image Classification | Tiny-ImageNet Bucket 1 Forget | Accuracy81.2 | 15 |