If you mean electron hopping between Fe2+ and Fe3+ (or another pair of metal cations), I can give a simple phenomenological explanation. Because of the nature of the superexchange interactions in Fe2+-O2--Fe3+ bonds in ferrites (also, cations should be in one crystallographic position, tethraedral or octaedral), exchanged electrons are hoping between cations, and cations change their valencies. We can't say its dipoles, but under an electric field, this hopping becomes directional. All charges that can alter the electric field contribute to polarization in delectrics. But when the frequency of altering the electric field becomes higher than the frequency of hoppin (or relaxation time of this process), hopping doesn't contribute to polarization.
here some references
Article Dielectric behaviour of Zn substituted Cu nano-ferrites
Article Electrophysical Characteristics of Polyvinyl Alcohol/Mn–Zn F...