Image Classification on model-vs-human OOD Aggregate (test)
80OOD AccuracyViT-22B-384
Evaluation results
| Method | Links | |
|---|---|---|
| ViT-22B-384Backbone=ViT-22B, Resolution=3842026.02 | 80 | |
| OpenCLIP ViT-H-14Backbone=ViT-H-142026.02 | 78 | |
| OpenCLIP ViT-H-14 Resized (64x64)Backbone=ViT-H-14, Input Transformation=Resized (64x64)2026.02 | 75 | |
| OpenCLIP ViT-H-14 Fourier-filteredBackbone=ViT-H-14, Input Transformation=Fourier-filtered2026.02 | 73 | |
| Humans (avg)Subject=Human2026.02 | 72 | |
| OpenCLIP ViT-H-14 BlurredBackbone=ViT-H-14, Input Transformation=Blurred (sigma = 2.5)2026.02 | 72 | |
| ImagenBackbone=Imagen2026.02 | 71 |