Classification on Derm7pt (test)
90.1AUCCNN MobileNet-V2
Evaluation results
| Method | Links | ||
|---|---|---|---|
| CNN MobileNet-V2Backbone=MobileNet-V2, Concept Bottleneck=None2026.07 | 90.1 | 84.8 | |
| G-CBM ResNet-50Backbone=ResNet-50, Concept Bottleneck=Interpretable concept-graph2026.07 | 86.8 | 82.3 | |
| CNN DenseNet-201Backbone=DenseNet-201, Concept Bottleneck=None2026.07 | 85 | 79.6 | |
| G-CBM DenseNet-201Backbone=DenseNet-201, Concept Bottleneck=Interpretable concept-graph2026.07 | 83.4 | 78.6 | |
| G-CBM MobileNet-V2Backbone=MobileNet-V2, Concept Bottleneck=Interpretable concept-graph2026.07 | 83 | 79.1 | |
| CNN ResNet-50Backbone=ResNet-50, Concept Bottleneck=None2026.07 | 82.8 | 79.1 | |
| MLP-CBM ResNet-50Backbone=ResNet-50, Concept Bottleneck=MLP2026.07 | 81 | 76.2 | |
| Linear-CBM ResNet-50Backbone=ResNet-50, Concept Bottleneck=Linear2026.07 | 81 | 72.4 |