How can a hidden layer optimize input features combination?

Hello William @spaceking ,

Please check this discussion out and in which you will find another link to an example of how neurons behave differently with respect to the same set of input features or input data. Let me know if you have a different concern or if you have other question.