Spread Complexity in Non-Hermitian Many-Body Localization Transition

TL;DR

This paper investigates how spread complexity metrics can distinguish phases and transitions in non-Hermitian many-body localization models, revealing differences based on time-reversal symmetry and boundary conditions.

Contribution

It introduces the use of spread complexity to detect ergodic, MBL, and real-complex eigenvalue transitions in non-Hermitian systems, highlighting differences due to TRS and boundary conditions.

Findings

Spread complexity distinguishes ergodic and MBL phases.

Thermofield double state complexity detects real-complex eigenvalue transition.

Charge density wave complexity varies with TRS in MBL phase.

Abstract

We study the behavior of spread complexity in the context of non-Hermitian many-body localization Transition (MBLT). Our analysis has shown that the singular value spread complexity is capable of distinguishing the ergodic and many-body localization (MBL) phase from the presaturation peak height for the non-hermitian models having time-reversal symmetry (TRS) and without TRS. On the other hand, the saturation value of the thermofield double (TFD) state complexity can detect the real-complex transition of the eigenvalues on increasing disorder strength. From the saturation value, we also distinguish the model with TRS and without TRS. The charge density wave complexity shows lower saturation values in the MBL phase for the model with TRS. However, the model without TRS shows a completely different behavior, which is also captivated by our analysis. So, our investigation unravels the real-complex transition in the eigenvalues, the difference between the model having TRS and without TRS, and the effect of boundary conditions for the non-hermitian models having MBL transitions, from the Krylov spread complexity perspective.