I have a question about the association between labels. Say my neural network performs multi-labeling in its output layer. Now, if one of the labels is for whether a person lives in city $X$, another for if they have a wiki article, and finally a third for if that person is a millionaire. Let us assume that these are highly correlated.

Is there some way for us to know if the neural network internally finds an association between these three labels?

  • $\begingroup$ As far as I understand, no. Neural networks are not explainable AI. Explainable AI is more based on symbolic AI (expert systems, inference rules, etc....) like RefPerSys hopes to become in 2022 $\endgroup$ Dec 11, 2021 at 18:17


You must log in to answer this question.