A Sudakov--Fernique proof of Lehner-type edge bounds for matrix-valued GUE sums
Abstract
Let A0,A1,…,An∈ MN(C) be Hermitian matrices and let G1,…,Gn be independent M× M GUE matrices normalized so that \|M-1/2Gi\| 2 almost surely as M∞. We study the spectral edges and operator norm of HM = A0 IM + 1MΣi=1n Ai Gi. Lehner's formula identifies the right and left edges of the corresponding free semicircular operator as ρ+ = ∈fZ 0λ(A0+Z+Σi=1n AiZ-1Ai) and ρ- = Z 0λ(A0+Z+Σi=1n AiZ-1Ai). Assuming Ai 0 for i 1 and M N, we prove via concentration and minimax duality the finite-dimensional bounds Eλ(HM) ρ+ + 9nN/M\,\|Σi=1n Ai2\|op1/2 and Eλ(HM) ρ- - 9nN/M\,\|Σi=1n Ai2\|op1/2. With ρ* = \ρ+,-ρ-\, this yields E\|HM\|op ρ* + 9nN/M\,\|Σi=1n Ai2\|op1/2. For uniformly bounded positive coefficients, bounded n, and N=o(M), one obtains M∞E\|HM\|opρ whenever ρ*,Mρ. The proof is a matrix-coefficient extension of classical Sudakov--Fernique comparison, combined with a Davidson--Szarek-type singular-value estimate and dual variational formulas for Lehner's edge quantities over density matrices. We also explain why this approach does not extend sharply to signed Hermitian coefficients.
Turn this paper into a full lesson
ArcXiv compiles a staged curriculum from this paper: 8-12 lessons across beginner → advanced, synthesised section guides, visuals, flashcards, a quiz, exercises, and on-demand deep dives per section. Grounded in the abstract, never invented.