Integrability versus chaos in the steady state of many-body open quantum systems
Josef Richter, Lucas S\'a, Masudul Haque

TL;DR
This paper investigates the distinction between integrable and chaotic steady states in many-body open quantum systems, revealing that both can have complex structures, but their operator-size distributions can serve as effective discriminators.
Contribution
It introduces a method to differentiate Liouvillian and steady-state chaos using level statistics and operator expansion analysis in open quantum systems.
Findings
Both chaotic and integrable steady states contain contributions from long-range and many-body operators.
Operator-size distribution effectively distinguishes between chaotic and integrable steady states.
Steady states are more complex than expected, involving many-body correlations regardless of integrability.
Abstract
The Lindblad description of an open quantum system gives rise to two types of integrability, since the nonequilibrium steady state can be integrable independently of the Liouvillian. Taking boundary-driven and dephasing spin chains as a representative example, we discriminate Liouvillian and steady-state chaos by combining level spacing statistics and an extension of the eigenstate thermalization hypothesis to open quantum systems. Moreover, we analyze the structure of the steady states by expanding it in the basis of Pauli strings and comparing the weight of strings of different lengths. We show that the natural expectation that integrable steady states are "simple" (i.e., built from few-body local operators) does not hold: the steady states of both chaotic and integrable models have relevant contributions coming from Pauli strings of all possible lengths, including long-range and…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
