Anyone want to do interpretable research on deep learning? Let's add a friend to have a detailed chat.