[0xDEAFBEAD]

I think this distinction is not sharp. You do hear ML practitioners talk about interpretability a lot.