Annotator Evaluation on CIFAR-10 (test)
78.32Test AccuracyGBV
Evaluation results
| Method | Links | |
|---|---|---|
| GBVBackbone=ResNet-18, Pre-trained weights=ImageNet, tau parameter setting=best τ2026.04 | 78.32 | |
| GBVBackbone=ResNet-18, Pre-trained weights=ImageNet, tau parameter setting=quick τ2026.04 | 77.94 | |
| MMDBackbone=ResNet-18, Pre-trained weights=ImageNet2026.04 | 75.95 | |
| DAVINZBackbone=ResNet-18, Pre-trained weights=ImageNet2026.04 | 74.82 | |
| LAVABackbone=ResNet-18, Pre-trained weights=ImageNet2026.04 | 74.74 | |
| UniformBackbone=ResNet-18, Pre-trained weights=ImageNet2026.04 | 74.37 |