This repository contains the code for the paper "SPADE: Sparsity Guided Debugging for Deep Neural Networks". In this paper we improved interpretability methods by first pruning the model on the sample ...