Image Classification on CIFAR-10 (test) (Accuracy and Delta)
91.8Test AccuracyTwin
Evaluation results
| Method | Links | ||
|---|---|---|---|
| TwinModel Class=ConvNet, Model=RN20, # Trials=1002024.03 | 91.8 | 0.2 | |
| m-sharpnessModel Class=ConvNet, Model=RN20, # Trials=1002024.03 | 91.6 | — | |
| TwinModel Class=ConvNet, Model=RNX11, # Trials=1002024.03 | 90.2 | 0.2 | |
| m-sharpnessModel Class=ConvNet, Model=RNX11, # Trials=1002024.03 | 90 | — | |
| TwinModel Class=ViT, Model=CCT-2/3x2, # Trials=492024.03 | 87.3 | 5.3 | |
| m-sharpnessModel Class=ViT, Model=CCT-2/3x2, # Trials=492024.03 | 82 | — |