English

Eigenvector decorrelation for random matrices

Probability 2026-03-03 v3 Mathematical Physics math.MP

Abstract

We study the sensitivity of the eigenvectors of random matrices, showing that even small perturbations make the eigenvectors almost orthogonal. More precisely, we consider two deformed Wigner matrices W+D1W+D_1, W+D2W+D_2 and show that their bulk eigenvectors become asymptotically orthogonal as soon as Tr(D1D2)21\mathrm{Tr}(D_1-D_2)^2\gg 1, or their respective energies are separated on a scale much bigger than the local eigenvalue spacing. Furthermore, we show that quadratic forms of eigenvectors of W+D1W+D_1, W+D2W+D_2 with any deterministic matrix ACN×NA\in\mathbf{C}^{N\times N} in a specific subspace of codimension one are of size N1/2N^{-1/2}. This proves a generalization of the Eigenstate Thermalization Hypothesis to eigenvectors belonging to two different spectral families.

Keywords

Cite

@article{arxiv.2410.10718,
  title  = {Eigenvector decorrelation for random matrices},
  author = {Giorgio Cipolloni and László Erdős and Joscha Henheik and Oleksii Kolupaiev},
  journal= {arXiv preprint arXiv:2410.10718},
  year   = {2026}
}

Comments

49 pages, 1 figure; v1 -> v2 -> v3: minor updates

R2 v1 2026-06-28T19:20:56.923Z