2015
DOI: 10.1088/1367-2630/17/12/122002
|View full text |Cite
|
Sign up to set email alerts
|

A random matrix model with localization and ergodic transitions

Abstract: Motivated by the problem of many-body localization and the recent numerical results for the level and eigenfunction statistics on the random regular graphs, a generalization of the Rosenzweig-Porter random matrix model is suggested that possesses two transitions. One of them is the Anderson localization transition from the localized to the extended states. The other one is the ergodic transition from the extended non-ergodic (multifractal) states to the extended ergodic states. We confirm the existence of both… Show more

Help me understand this report
View preprint versions

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

36
423
4

Year Published

2016
2016
2023
2023

Publication Types

Select...
7

Relationship

0
7

Authors

Journals

citations
Cited by 216 publications
(463 citation statements)
references
References 39 publications
(102 reference statements)
36
423
4
Order By: Relevance
“…Note that this linear spectrum for the Localized phase has already been found for the MBL case in [113] and for an Anderson Localization matrix model in [53].…”
Section: Many-body-localized Phasesupporting
confidence: 74%
See 2 more Smart Citations
“…Note that this linear spectrum for the Localized phase has already been found for the MBL case in [113] and for an Anderson Localization matrix model in [53].…”
Section: Many-body-localized Phasesupporting
confidence: 74%
“…36 (as found for Anderson localization on trees [51,112] or as in some matrix model [53]). One general constraint is however that the minimal exponent has to be strictly positive…”
Section: Fmentioning
confidence: 56%
See 1 more Smart Citation
“…Despite this fact, integrable matrices do exhibit a parameter-dependent localization property [43] in which almost all eigenstates of the matrix H(x) = xT + V are localized in the basis of V for all values of x. The stability of this property when a random matrix perturbation is added to H(x), including the possibility of a multifractal phase accompanying the localized and delocalized regimes [40], is the subject of future study.…”
Section: Discussionmentioning
confidence: 98%
“…(11) and Refs. 24, 25. It is interesting to note that many-body localized [36] (MBL) systems are also expected to display Poisson level statistics [37,38], and there exist random matrix ensembles which model localization and its statistical signatures [39,40]. Such ensembles are basis-dependent, which is natural because localization is a basis-dependent property.…”
Section: Discussionmentioning
confidence: 99%