Probing ViTs

non-profit

AI & ML interests

We are interested to study the representations learned by Vision Transformers.

Organization Card

About org cards

DINO	Dog

spaces 4

Attention Heat Maps

Class Attention Map

Class Saliency Map

Attention Rollout

models 4

probing-vits/cait_xxs24_224_classification

Updated May 2, 2022 • 9 • 3

probing-vits/vit_b16_patch16_224_i1k

Updated Apr 18, 2022 • 10

probing-vits/vit-dino-base16

Updated Apr 12, 2022 • 14 • 3

probing-vits/vit_b16_patch16_224_i21k_i1k

Updated Apr 12, 2022 • 8

datasets

None public yet