Lower Bound on Locally Accessible Information: Local Subentropy
Apratim De, Sen U., Maciej Lewenstein
Abstract
We find a universal lower bound on locally accessible information for arbitrary bipartite ensembles, when one of the parties is two-dimensional. In higher dimensions and in higher number of parties, the lower bound is on accessible information by separable operations. We evaluate this bound for several ensembles, for which the exact locally accessible information is not known. We show that for any given density matrix (of arbitrary number of parties and dimensions), there exists an ensemble, the so-called ``Scrooge ensemble'', which averages to the given density matrix and whose locally accessible information saturates the lower bound.