Enhancing Pre-trained Representation Classifiability can Boost its Interpretability.

Shufan Shen, Zhaobo Qi, Junshu Sun, Qingming Huang, Qi Tian, Shuhui Wang

Browse the full ICLR paper archive.