Training Neural Networks for Modularity aids Interpretability

Open in new window