This equivalence shows that neural networks are indeed interpretable by design and makes the \textit{black-box} understanding obsolete Show me the decision tre

This equivalence shows that neural networks are indeed interpretable by design and makes the \textit{black-box} understanding obsolete Show me the decision tree for a big classifier trained on imagenet, and let's see if it's interoperable