Is there a sense in which classical simulability and Hamiltonian learning should be linked? I often think one as being the inverse of the other (exp values from couplings, vs, couplings from exp values) but this seems naive, and likely proven wrong by
11 months ago