Fetching the paper…

SelfExplain: A Self-Explaining Architecture for Neural Text Classifiers · Around