The Inside of a Model - Part C: Why MLPs Were Not Enough

Up until now, we talked about MLPs, how they process vectors, and how they learn using backpropagation. We saw that a neural network can take an input vector, pass it through layers, calculate a loss,