Café com Física
Shared information at criticality. Equilibrium and nonequilibrium stationary state
We consider bipartitions of classical statistical mechanical systems either at equilibrium or describing the stationary states of stochastic models. Inspired by information theory we define estimators of the shared information between the two subsystems. As an application we compute the estimators for a variety of one-dimensional stationary states. They are all finite if the correlation lengths are finite and diverge if the systems are critical. Their dependence on the size of the system is the same for each universality class. If the stochastic process is conformal invariant, the estimators behave similarly to the entanglement entropy of the ground-state wavefunction describing a quantum phase transition.