Universality
A. B. J. Kuijlaars
TL;DR
This work formalizes universality in random matrix theory by showing that local eigenvalue statistics converge, under appropriate centering and scaling, to universal kernels that depend only on symmetry class and spectral location. For unitary ensembles, Local statistics are determinantal with limiting kernels such as $K^{\sin}$ in the bulk, $K^{\mathrm{Ai}}$ at soft edges, and $K^{\mathrm{Bes},\alpha}$ at hard edges; orthogonal and symplectic ensembles yield Pfaffian analogs $K^{\mathrm{Ai},\beta}$ and $K^{\sin,\beta}$. The Riemann–Hilbert method, via the Deift–Zhou steepest descent, provides a robust framework to derive these limits in the one-cut regular case and to obtain universal asymptotics for orthogonal polynomials and associated kernels. The paper also surveys nonstandard universality at spectrum singularities, introducing Painlevé II and I hierarchy kernels and Pearcey kernels for interior, edge, and exterior singular points. Together, these results illuminate how microscopic spectral fluctuations are governed by universal structures, with broad implications for statistical physics, number theory, and beyond.
Abstract
Universality of eigenvalue spacings is one of the basic characteristics of random matrices. We give the precise meaning of universality and discuss the standard universality classes (sine, Airy, Bessel) and their appearance in unitary, orthogonal, and symplectic ensembles. The Riemann-Hilbert problem for orthogonal polynomials is one possible tool to derive universality in unitary random matrix ensembles. An overview is presented of the Deift/Zhou steepest descent analysis of the Riemann-Hilbert problem in the one-cut regular case. Non-standard universality classes that arise at singular points in the spectrum are discussed at the end.