CNN Explainer is a useful tool to interactively learn how CNNs for image classification work behind the scenes. It runs a pre-trained small VGG model in the browser using TensorFlow.js. Common terms such as activation functions, the different kinds of layers and hyperparamters (i.e. padding, kernel size, stride) and their interactions are explained. It is even possible to upload an image and see how it is classified. With this tool the authors shed some light into the classification process of Convolutional Neural Networks that are often labeled as black boxes.