Image Classification on ImageNet 1K (val) (kNN and Linear Evaluation Metrics)
78.66kNN AccuracyiBOT-vMF
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| iBOT-vMFArchitecture=ViT-Base/162024.05 | 78.66 | 80.2 | 80.27 | — | |
| DINO-vMFArchitecture=ViT-Base/162024.05 | 77.4 | 78.67 | 78.81 | — | |
| STELLAR# TKS=16, Backbone=ViT-H2026.02 | 77.31 | — | — | 79.1 | |
| iBOTArchitecture=ViT-Base/162024.05 | 77.1 | 79.33 | 79.5 | — | |
| DINOArchitecture=ViT-Base/162024.05 | 76.11 | 77.88 | 78.17 | — | |
| DINO# TKS=1, Backbone=ViT-B, Token=CLS2026.02 | 74.69 | — | — | 76.46 | |
| MSNArchitecture=ViT-Base/162024.05 | 73.33 | 75.84 | 74.8 | — | |
| STELLAR# TKS=16, Backbone=ViT-B2026.02 | 67.25 | — | — | 73.26 | |
| STELLAR# TKS=196, Backbone=ViT-B2026.02 | 64.71 | — | — | 72.21 | |
| DINO# TKS=196, Backbone=ViT-B2026.02 | 54.41 | — | — | 70.31 | |
| MAE# TKS=196, Backbone=ViT-B2026.02 | 25.82 | — | — | 66.32 | |
| TiTok# TKS=32, Backbone=ViT-B, Decoder=Native ViT decoder of larger size2026.02 | 7.3 | — | — | 33.42 | |
| TiTok# TKS=64, Backbone=ViT-B, Decoder=Native ViT decoder of larger size2026.02 | 7.29 | — | — | 32.87 | |
| BeITArchitecture=ViT-Base/162024.05 | — | — | 56.7 | — | |
| BeIT-v2Architecture=ViT-Base/162024.05 | — | — | 80.1 | — | |
| MAEArchitecture=ViT-Base/162024.05 | — | — | 68 | — | |
| MoCo-v3Architecture=ViT-Base/162024.05 | — | — | 76.7 | — |