| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Medical Image Classification | PathMNIST | Accuracy99.38 | 61 | |
| Image Classification | PathMNIST | Clean Accuracy89.45 | 60 | |
| Medical Image Classification | PATHMNIST | Clean Accuracy89.45 | 48 | |
| Image Classification | PathMNIST v1 (test) | Accuracy92.02 | 36 | |
| Image Classification | PathMNIST v2 (test) | Accuracy89.86 | 35 | |
| Image Classification | PathMNIST | Accuracy96.65 | 34 | |
| 9-class classification | PathMNIST | Accuracy93.73 | 32 | |
| Medical Image Classification | PathMNIST v2 (test) | Accuracy97.5 | 22 | |
| Medical Image Classification | PathMNIST 40% Noise | Sensitivity96.8 | 10 | |
| Medical Image Classification | PathMNIST 20% Noise | Sensitivity95 | 10 | |
| Medical Image Classification | PathMNIST 0% Noise | Sensitivity88.7 | 10 | |
| Machine Unlearning | PathMNIST External (test) | Residual Accuracy (Acc_r)88.63 | 9 | |
| Machine Unlearning | PathMNIST Internal (test) | Accuracy (Retained)97.78 | 9 | |
| Classification | PathMNIST D11 | PGD Acc45.7 | 7 | |
| Image Classification | PathMNIST | BP Top-1 Accuracy88.89 | 6 | |
| Federated Unlearning | PathMNIST | Training Time (s)350 | 6 | |
| Image Classification | PathMNIST-C | Accuracy52 | 6 | |
| Conformal Prediction | PathMNIST (test) | Coverage95 | 5 | |
| Machine Unlearning | PathMNIST 50% removal binary (test) | Specificity99 | 5 | |
| Machine Unlearning | PathMNIST 20% removal binary (test) | Specificity99 | 5 | |
| Failure types identification | BP-MNIST | Learnability92 | 5 | |
| Image Classification | PATHMNIST MedMNIST (test) | Clean Accuracy84.88 | 4 | |
| Image Classification | PathMNIST (test) | Accuracy@1 (%)78.88 | 4 | |
| Learning to Defer | PathMNIST Specialist | Accuracy87.74 | 2 | |
| Learning to Defer | PathMNIST Clean | Accuracy87.31 | 2 |