Information in the Weights and Emergent Properties of Deep Neural Networks Part II