| semi-supervised-image-classification-on-imagenet-1-labeled-data | MSN (ViT-B/4) | #9 | Top 1 Accuracy: 75.7% |
| self-supervised-image-classification-on-imagenet | MSN (ViT-L/7) | #20 | Top 1 Accuracy: 80.7%Number of Params: 306M |
| image-classification-on-cifar-10 | MSN ViT-B/16 | #32 | Percentage correct: 99.0 |
| image-classification-on-cifar-10 | MSN ViT-B/16 (linear probe) | #151 | Percentage correct: 95.7 |
| image-classification-on-cifar-100 | MSN ViT-B/16 | #36 | Percentage correct: 90.5 |
| image-classification-on-cifar-100 | MSN ViT-B/16 (linear probe) | #124 | Percentage correct: 82.8 |
| image-classification-on-imagenet | MSN ViT-B/16 | #461 | Top 1 Accuracy: 83.4% |
| image-classification-on-imagenet | MSN ViT-L/7 | #724 | Top 1 Accuracy: 80.7% |