The output from the convolutional layer is often handed from the ReLU activation functionality to bring non-linearity on the model. It takes the characteristic map and replaces all of the adverse values with zero. A VGG-block had lots of 3x3 convolutions padded by one to keep the size of https://financefeeds.com/5-best-copyright-to-buy-now-heres-why-this-layer-1-blockchain-presale-is-winning-hearts/