Three non-Hermitian random matrix universality classes of complex edge statistics: Spacing ratios and distributions
Gernot Akemann, Georg Angermann, Noah Aygün, Adam Mielke, · 2026
Read Three non-Hermitian random matrix universality classes of complex edge statistics: Spacing ratios and distributions by Gernot Akemann, Georg Angermann, Noah Aygün, Adam Mielke, (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-28457.txt.
Introduction
Akemann and six co-authors compare local statistics at the edges of three non-Hermitian random matrix classes. They study complex spacing ratios and nearest-neighbor gaps for the Ginibre, symmetric, and self-dual ensembles. In the first class they clean up exact finite-size formulas and test a three-by-three guess that bends toward the unitary case. Numerical work across all three classes matches a picture of charges repelling in a two-dimensional Coulomb gas. The spacing ratio alone, they find, misses some edge detail.
Catalog record
| CREATOR | Gernot Akemann, Georg Angermann, Noah Aygün, Adam Mielke, |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 113743 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.28457
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Three non-Hermitian random matrix universality classes of complex edge statistics: Spacing ratios and distributions Gernot Akemann, Georg Angermann, Noah Aygün, Adam Mielke, Abstract The conjectured three generic local bulk statistics amongst all non-Hermitian random matrix symmetry classes have recently been extended to three generic local edge statistics. We study analytically and numerically complex spacing ratios and nearest-neighbour (NN) spacing distributions that characterise such local statistics. We choose the three simplest representatives of these universality classes, given by the Gaussian ensembles of complex Ginibre, complex symmetric and complex self-dual matrices, denoted by class A, AI † and AII † . In the first part, we analytically study the complex spacing ratio in class A, at finite matrix size N N . Introducing a conditional point process, we simplify existing expressions and show why an uncontrolled approximation introduced earlier converges well in the large- N N limit in the bulk. When specifying to the elliptic Ginibre ensemble, we present a parameter-dependent N = 3 N=3 surmise for the complex spacing ratio, interpolating to that of the Gaussian unitary ensemble (GUE), where such a surmise is very accurate. In the second numerical part, we compare complex spacing ratios, its moments, and NN spacing distributions for all three ensembles with that of uncorrelated points, the two-dimensional (2D) Poisson process, both in the bulk and at the edge. The varying degree of repulsion within these different edge universality classes can be well understood in terms of an effective 2D Coulomb gas description, at different values of inverse temperature β \beta . We find indications that the complex spacing ratio does not fully unfold the local statistics at the edge. Finally we verify that for small argument, in all three symmetry classes the NN spacing distributions in the bulk and at the edge are consistent with the universal cubic repulsion. Dedicated to the memory of Santosh Kumar 1 Introduction The application of non-Hermitian random matrix theory (RMT) to physics is a very active area of research. This was initiated by the generalisation of the quantum chaos conjecture to dissipative open quantum chaotic system [ 1 ] , and nowadays goes much beyond this area of research, see [ 2 ] for a review. Whilst in many applications the description of data by RMT is heuristic, there are examples for a precise map to an RMT description, where we will focus on non-Hermitian examples only. Such a link exists for non-interacting Fermions in 2D in a rotating trap, or equivalently perpendicular magnetic field, where the ground state properties can be mapped to the complex Ginibre ensemble (GinUE) [ 3 , 4 , 5 ] . Extensions beyond the lowest Landau levels or zero temperature exist, going beyond standard RMT classes, cf. [ 6 ] and references therein. A second example relates non-Hermitian RMT to effective field theory of Quantum Chromodynamics, with finite chemical potential [ 7 ] or finite lattice spacing [ 8 ] . The corresponding symmetry classes are the chiral partners of the complex elliptic Ginibre ensemble [ 7 ] , and the Wilson–Dirac class [ 9 ] . The application to dissipative open quantum chaotic systems [ 1 ] has seen a strong revival recently, with many authors contributing, including very prominently Santosh Kumar and his group. These works include the dissipative spectral form factor [ 10 , 11 ] , the number variance [ 12 ] , the complex spacing ratio [ 13 , 14 , 15 ] introduced in [ 16 ] , and the NN spacing distribution in radial distance in the complex plane [ 17 ] as tools, with the latter two being local. They are applied to physical systems such as the kicked top or rotor [ 1 , 12 , 13 , 15 ] and in open dissipative quantum spin chains [ 18 , 19 ] to name few examples. The main focus is on the transition between integrable and chaotic open quantum systems, characterised by 2D Poisson respectively random matrix statistics in the corresponding symmetry class—the content of the quantum chaos conjecture. The study of symmetry classes in non-Hermitian RMT has been revisited recently. In 2019 Kawabata et al. [ 20 ] reanalysed the classification of Bernard–LeClair [ 21 ] and Magnea [ 22 ] from 2002, finding a total number of 38 symmetry classes. 1 1 1 A revised version of [ 21 ] was published after the work [ 20 ] appeared, it agrees with this total number of classes. Shortly thereafter, it was proposed by Hamazaki et al. [ 19 ] , based on numerics of the NN spacing distribution and heuristic arguments, that only 3 generic local bulk universality classes exist amongst all symmetry classes. This conjecture was later extended to the spectral edge [ 23 ] . For comparison, in the case of Hermitian random matrices, where 10 symmetry classes are identified [ 24 ] , only three different local bulk and three different local edge statistics are known which are generic. These can be represented by the three Dyson-ensembles [ 25 ] . In the non-Hermitian situation, the three simplest representatives are the classes A (complex Ginibre), AI † (complex symmetric) and AII † (complex self-dual) which are all Gaussian, as put forward by [ 19 ] . Much earlier, it has been conjectured on simple arguments from level crossings in quantum mechanical systems, that in the small argument expansion of the different NN spacing distributions all show a cubic vanishing that is universal [ 26 , 27 ] . What is known analytically about the NN spacing distribution and the complex spacing ratio in the three symmetry classes A, AI † and AII † ? The complex eigenvalues in class A represent a determinantal point process, which is well described by the universal exponential Ginibre kernel in the bulk, and the complementary error function kernel (also called Faddeeva plasma kernel) at the edge of the spectrum, both known at finite N and in the limit of infinite matrix dimension N\to\infty . Closed form expressions exist for the gap probabilities in terms of the Fredholm determinant of these two kernels, which are most explicitly known in the bulk [ 1 ] . This leads to explicit expressions for the NN and next-to-nearest neighbour (NNN) spacing distribution [ 1 ] . For a recent review on the Ginibre and related ensembles we refer to [ 28 ] . In contrast, even in class A only approximate expressions exist for the complex spacing ratio [ 16 , 29 ] . They include an N -dependent determinantal expression and a surmise at N=3 , which is not very accurate. This is in stark contrast to the spacing ratio in Hermitian RMT introduced by [ 30 , 31 ] in the bulk of the real spectrum, which is very well approximated by N=3 . The idea behind the spacing ratio is that unfolding (see e.g. [ 27 ] ), that is necessary to compare local statistics from data to RMT, drops out in taking the ratio of the difference to the NN and NNN eigenvalue. Exact analytic limiting results are scarce for the spacing ratio, see however [ 32 , 33 ] . Higher order