Formal Local Implication Between Two Neural Networks
Given two neural network classifiers with the same in-put and output domains, our goal is to compare the two networksin relation to each other over an entire input region (e.g., within avicinity of an input sample). To this end, we establish the foundationof formal local implication between two networks, i.e., N2D=⇒ N1, in an entire input region D. That is, network N1 consistently makesa correct d