Salient ImageNet: How to discover spurious features in Deep Learning?

we identify spurious or core neural features (penultimate layer neurons of a robust model) via limited human supervision (e.g., using top 5 activating images per feature). We then show that these neural feature annotations generalize extremely well to many more images without any human supervision. We use the activation maps for these neural features as the soft masks to highlight spurious or core visual features. Using this methodology, we introduce the Salient Imagenet dataset containing core and spurious masks for a large set of samples from Imagenet. Using this dataset, we show that several popular Imagenet models rely heavily on various spurious features in their predictions, in-dicating the standard accuracy alone is not sufficient to fully assess model perfor-1