The input layer is where we feed input to the network. The number of neurons in the input layer is the number of inputs we feed to the network. Each input will have some influence on predicting the output and this will be multiplied by weights, while bias will be added and passed to the next layer.