Many-body localization landscape
Shankar Balasubramanian, Yunxiang Liao, Victor Galitski
Abstract
Recently, a structure called the localization landscape was shown to accurately predict the location of eigenfunctions in the single-particle Anderson model. Here, the authors generalize the localization landscape to a large class of interacting many-body systems. They use the many-body localization landscape (MBLL) to prove decay estimates of eigenstates, and show robustness of the MBLL to weak interactions and hopping by deriving a convergent locator expansion. This establishes the MBLL as a method for efficiently estimating low-energy eigenstates and understanding possible phases in MBL models.
Topics & Concepts
EigenfunctionRobustness (evolution)Eigenvalues and eigenvectorsEnergy landscapeStatistical physicsComputer sciencePhysicsBiologyQuantum mechanicsThermodynamicsGeneBiochemistryQuantum many-body systemsPhysics of Superconductivity and MagnetismTensor decomposition and applications