| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Image Classification | FashionMNIST (test) | Accuracy94.83 | 218 | |
| Image Classification | FashionMNIST | Accuracy96.93 | 147 | |
| Anomaly Detection | FashionMNIST (test) | ROCAUC0.959 | 35 | |
| Poisoning Attack Mitigation | FashionMNIST | F1 Score87.8 | 19 | |
| Out-of-Distribution Detection | FashionMNIST (test) | AUROC0.917 | 14 | |
| Image Reconstruction | FashionMNIST (test) | MSE0.403 | 12 | |
| Deep Active Learning | FashionMNIST | AUBC0.7812 | 12 | |
| Image modeling | FashionMNIST (test) | FID52 | 12 | |
| Fine-grained accuracy control attack | FashionMNIST | Control Error Rate0.59 | 10 | |
| Image Classification | FashionMNIST OOD: MNIST (test) | AUROC (H)0.971 | 10 | |
| Denoising Autoencoding | FashionMNIST | PSNR24.91 | 10 | |
| Novelty Detection | FashionMNIST [0,2,3,7,8] (ID) vs [1,4,5,6,9] (OOD) (unlabeled set) | AUROC95 | 10 | |
| Image Classification | FashionMNIST IDN-50% | Accuracy78.19 | 9 | |
| Image Classification | FashionMNIST IDN-45% | Accuracy88.75 | 9 | |
| Image Classification | FashionMNIST IDN-40% | Accuracy90.01 | 9 | |
| Image Classification | FashionMNIST IDN-30% | Accuracy90.68 | 9 | |
| Image Classification | FashionMNIST IDN-20% | Accuracy91.58 | 9 | |
| Classification | FashionMNIST linf, epsilon=0.3 (test) | Robust Accuracy (FGSM)96.34 | 8 | |
| Classification | FashionMNIST (l2, epsilon=2) (test) | PGD (100) Robust Accuracy82.91 | 8 | |
| Classification | FashionMNIST l1, epsilon=10 (test) | Clean Accuracy92.56 | 8 | |
| Classification | FASHIONMNIST | Empirical Risk12.6 | 8 | |
| Out-of-Distribution Detection | FashionMNIST (ID) vs Rotation (avg) (OoD) | AUROC79 | 8 | |
| Out-of-Distribution Detection | FASHIONMNIST vs Rotation (test) | AUROC0.9 | 8 | |
| Image Classification | FashionMNIST-10 (test) | Test Accuracy72.11 | 6 | |
| Federated Learning | FashionMNIST | Computational Time (s)0.019 | 5 |