The card describes this checkpoint as ViT from Dosovitskiy et al. but its first BibTeX entry currently points to a different paper. This swaps that entry for the citation published by the official Vision Transformer repository. The ImageNet citation remains unchanged.
