Learning Variational Word Masks to Improve the Interpretability of Neural Text Classifiers.

Hanjie Chen, Yangfeng Ji

Browse the full EMNLP paper archive.