Mathematics of Disordered Quantum Systems and Matrices
Mathematics
1 Career
- Since 2013 Professor, Institute of Science and Technology Austria (ISTA)
- 2003 – 2013 Professor (C4/W3) Ludwig-Maximilians University, Munich, Germany
- 2003 – 2004 Professor, Georgia Institute of Technology, Atlanta, USA
- 2001 – 2003 Associate Professor, Georgia Institute of Technology, Atlanta, USA
- 1998 – 2001 Assistant Professor, Georgia Institute of Technology, Atlanta, USA
- 1995 – 1998 Courant Instructor, New York University, New York, USA
- 1994 – 1995 Postdoctoral Fellow, ETH Zurich, Switzerland
1.1 Selected Distinctions
- 2026 Leroy P. Steele Prize for Seminal Contribution to Research of the American Mathematical Society
- 2025 Elected member of Leopoldina (German National Academy of Sciences)
- 2022 Fellow of the American Mathematical Society
- 2020 Erwin Schrödinger Prize of the Austrian Academy of Sciences
- 2017 Leonard Eisenbud Prize of the American Mathematical Society
- 2016 Elected Foreign Member of the Hungarian Academy of Sciences
- 2015 Elected Member of the Academia Europaea
- 2015 Elected Corresponding Member of the Austrian Academy of Sciences
- 2014 Invited Speaker at the International Congress of Mathematicians
1.2 Selected Grants
- 2021 – 2026 ERC Advanced Grant “RMTBEYOND” (No. 101020331), € 1,912K.
- 2014 – 2019 ERC Advanced Grant “RANMAT” (No. 338804), € 1,755K.
1.3 Tech Transfer (Patents, Spin-Offs)
N/A
1.4 Selected Publications
- G. Cipolloni, L. Erdős, D. Schröder, Eigenstate Thermalization Hypothesis for Wigner Matrices. Commun. Math. Phys. 388, 1005–1048 (2021).
- G. Cipolloni, L. Erdős, D. Schröder, Central limit theorem for linear eigenvalue statistics of non-Hermitian random matrices. Comm. Pure Appl. Math. 76(5), 899–1136 (2023).
- L. Erdős, H. C. Ji, Wegner estimate and upper bound on the eigenvalue condition number of non-Hermitian random matrices. Comm. Pure Appl. Math. 77(9), 3785–3840 (2024).
- L. Erdős, V. Riabov, Eigenstate Thermalization Hypothesis for Wigner-type Matrices. Commun. Math. Physics. 405 282 (2024).
- A. Campbell, G. Cipolloni, L. Erdős, Hong Chang Ji, On the spectral edge of non-Hermitian random matrices. Ann. Probab. 53, No. 6, 2256–2308 (2025)
2 Group Members
- Oleksii Kolupaiev, PhD Student since 2022
- Andrew Campbell, Postdoc since 2025
- Jaehun Lee, Postdoc since 2025
- Volodymyr Riabov, Postdoc since 2025
3 Research
Large random matrices are fundamental mathematical objects whose study is motivated by several applications, including statistics, disordered quantum systems, quantum information theory and more recently by neural networks. The richness of the area is best exemplified by Eugene Wigner’s revolutionary observation in 1950’s that the statistical properties of the energy spectra of large atomic nuclei follow universal patterns which can be effectively described by eigenvalues of random matrices. This was complemented by the discovery the Anderson phase transition in 1958 and the Bohigas-Giannoni-Schmitt conjecture from 1984 asserting that the spectra of disordered quantum systems in the chaotic regime follow random matrix behavior. Our group is devoted to the rigorous mathematical study of these phenomena. Realistic quantum systems are still largely out of reach, but a spectacular progress has been made in the last two decades on more and more realistic models. Most importantly, the Wigner-Dyson-Mehta conjecture on the universality of the local eigenvalue statistics was proven in 2008 for Wigner matrices. Our first main goal is to go beyond such homogeneous mean field models and develop tools for proving the universality conjecture for disordered quantum systems with nontrivial spatial structure. Our second main goal is to extend random matrix theory to non-Hermitian matrices.
3.1 Current Scientific Projects/Directions
- Eigenstate Thermalisation Hypothesis for general correlated matrices
- Gaussian fluctuation for linear eigenvalue statistics for general correlated matrices
- Hyperuniformity of eigenvalues of large non-Hermitian matrices with i.i.d. entries
- Eigenvalue correlations of large random matrices and their minors
- Roots of random polynomials
3.2 Research in Focus
Eigenstate Thermalisation Hypothesis (ETH) is a fundamental physical concept, also known as Quantum Unique Ergodicity (QUE) in the mathematics literature, that asserts that eigenfunctions of sufficiently chaotic quantum systems (often arising as quantisations of ergodic classical dynamical systems) are uniformly distributed in the phase space. More precisely, quantum expectations of general deterministic observables A converge to the microcanonical expectation of A as the system size increases, moreover they exhibit universal Gaussian fluctuations. This paradigm can be viewed as the eigenfunction analogue of Wigner’s universality vision for eigenvalues. In a series of works our group has developed a method, dubbed as the zigzag strategy, that, among many other results, proves ETH for a very large class of random matrices, in particular validating Deutsch’ conjecture from 1991. Concretely, in 2021 we proved ETH for Wigner matrices (item (1) in the list of selected publications) in 2024 for Wigner type matrices that involve nontrivial inhomogeneous structure (item (2)); in 2025 for one dimensional random band matrices (arxiv:2506.06441) and currently we are finishing the proof of ETH for mean field random matrices with general correlation structure.
