OPE Prop just got an addition in the form of a new formula that explains how to derive a calculation formula for any activation function with the MSE error function. It is available here.