Distance-independent characterization of asymmetric relaxation

Develop a framework to describe asymmetric relaxation that does not depend on the specific divergence or distance to equilibrium chosen on the thermodynamic state manifold (e.g., thermodynamic availability or Kullback–Leibler divergence), ensuring that asymmetry predictions are invariant under the choice of distance measure.

Background

The analysis in the paper measures distance to equilibrium via the thermodynamic availability (a Bregman divergence related to the Kullback–Leibler divergence), which underpins the gradient-flow representation and the asymmetry criteria.

The authors emphasize that availability is only one of many possible divergences on the state manifold and explicitly highlight the need for a description of asymmetric relaxation that does not hinge on any particular distance measure.