Image Classification on CIFAR-100 80% label flips (test)
17.52AccuracyNoTrans
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| NoTransBackbone=CNNs2025.05 | 17.52 | 0.8252 | 0.1663 | 0.6 | |
| REFINEBackbone=CNNs2025.05 | 17.37 | 0.8239 | 0.1641 | 0.2 | |
| DistillBackbone=CNNs2025.05 | 15.11 | 0.8174 | 0.1227 | 0 | |
| LoRABackbone=CNNs2025.05 | 2.01 | 0.6251 | 0.0026 | 0 | |
| LinearProbeBackbone=CNNs2025.05 | 1 | 0.674 | 0.0002 | 0 | |
| AdapterBackbone=CNNs2025.05 | 1 | 0.525 | 0.0002 | 0 | |
| DANN-GateBackbone=CNNs2025.05 | 1 | 0.5754 | 0.0002 | 0 |