| zero-shot-transfer-image-classification-on-imagenet-v2 | LiT-22B | #1 | Accuracy (Private): 80.9 |
| zero-shot-transfer-image-classification-on-objectnet | LiT-22B | #1 | Accuracy (Private): 87.6 |
| zero-shot-transfer-image-classification-on-imagenet | LiT-22B | #3 | Accuracy (Private): 85.9 |
| object-recognition-on-shape-bias | ViT-22B-384 | #4 | shape bias: 86.4 |
| object-recognition-on-shape-bias | ViT-22B-560 | #5 | shape bias: 83.8 |
| object-recognition-on-shape-bias | ViT-22B-224 | #7 | shape bias: 78.0 |
| zero-shot-transfer-image-classification-on-imagenet-a | LiT-22B | #4 | Accuracy (Private): 90.1 |
| zero-shot-transfer-image-classification-on-imagenet-r | LiT-22B | #4 | Accuracy: 96.0 |
| image-classification-on-imagenet | ViT-L/16 (384res, distilled from ViT-22B) | #19 | Top 1 Accuracy: 89.6%Number of params: 307M |
| image-classification-on-imagenet | ViT-B/16 | #42 | Top 1 Accuracy: 88.6%Number of params: 86M |
| action-classification-on-kinetics-400 | ViT-22B | #25 | Acc@1: 88.0 |