These exercises provide us with additional exposure to the parity problem. What happens to the learning of multilayer perceptrons when the problem is scaled up in size ? We will find out by training two different types of multilayer perceptrons on two different versions of the parity problem, 5-bit and 7-bit.