Our paper, Interpreting Internal Activation Patterns in Deep Temporal Neural Networks by Finding Prototypes, written by Sohee Cho, Wonjoon Chang, Ginkyeng Lee and Jaesik Choi is accepted at KDD-2021.