What I'm looking for is a non-asymptotic bound on the probability that the smallest gap between eigenvalues of a GUE matrix does not exceed a certain value.

I'm aware of the bounds in http://imrn.oxfordjournals.org/content/2010/3/436.full.pdf and https://people.math.osu.edu/nguyen.1261/cikk/gap.pdf , but was wondering if there is a short/elementary way to derive such bounds in the special case of a GUE.


This is studied in

Gérard Ben Arous and Paul Bourgade, Extreme gaps between eigenvalues of random matrices, Ann. Probab. 41 (2013), no. 4, 2648--2681.

(Ah, so that's how the "insert citation" button works!) In particular, for GUE, the smallest gap has size about $n^{-4/3}$, and after multiplying by $n^{4/3}$, converges to a distribution with pdf $3 x^2 e^{-x^3}$. I think in principle a quantitative rate of convergence can be extracted from the methods of the paper, though it does not appear to have been done explicitly there.

  • $\begingroup$ Thanks! I was aware of that paper (actually, I was just looking at it). I'll see if I can extract some quantitative bounds from that. $\endgroup$ – Martin Lotz Dec 11 '15 at 9:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.