Distilling the Knowledge in a Neural Network - Explained Simply | ArXiv Explained