Training Neural Networks for Modularity aids Interpretability

Re-assign community