Abstract
Large deviation behavior of the largest eigenvalue \(\lambda _1\) of Wigner matrices including those arising from an Erdős-Rényi random graph \({\mathcal {G}}_{n,p}\) with i.i.d. random conductances on the edges has been the topic of considerable interest. However, despite several recent advances, not much is known when the underlying graph is sparse i.e., \(p\rightarrow 0\), except the recent works (Bhattacharya et al., Ann Probab 49(4):1847–1885, 2021and Bhattacharya and Ganguly, SIAM J Discret Math, 2020) which consider the simpler case of the graph without additional edge weights. Under sufficiently general conditions on the conductance distribution, one expects the ‘dense’ behavior as long as the average degree np is at least logarithmic in n. In this article we focus on the case of constant average degree i.e., \(p=\frac{d}{n}\) for some fixed \(d>0\) with standard Gaussian weights. Results in Bandeira and Van Handel (Ann Probab 44(4):2479–2506, 2016) about general non-homogeneous Gaussian matrices imply that in this regime \(\lambda _1\) scales like \(\sqrt{\log n}.\) We prove the following results towards a precise understanding of the large deviation behavior in this setting.
-
1.
(Upper tail probabilities and structure theorem): For \(\delta >0,\) we pin down the exact exponent \(\psi (\delta )\) such that
$$\begin{aligned} {\mathbb {P}}(\lambda _1\ge \sqrt{2(1+\delta )\log n})=n^{-\psi (\delta )+o(1)}. \end{aligned}$$Further, we show that conditioned on the upper tail event, with high probability, a unique maximal clique emerges with a very precise \(\delta \) dependent size (takes either one or two possible values) and the Gaussian weights are uniformly high in absolute value on the edges in the clique. Finally, we also prove an optimal localization result for the leading eigenvector, showing that it allocates most of its mass on the aforementioned clique which is spread uniformly across its vertices.
-
2.
(Lower tail probabilities): The exact stretched exponential behavior
$$\begin{aligned} {\mathbb {P}}(\lambda _1\le \sqrt{2(1-\delta )\log n})=\exp \left( -n^{\ell (\delta )+o(1)}\right) \end{aligned}$$is also established.
As an immediate corollary, one obtains that \(\lambda _1\) is typically \((1+o(1))\sqrt{2\log n}\), a result which surprisingly appears to be new. A key ingredient in our proofs is an extremal spectral theory for weighted graphs obtained by an \(\ell _1-\)reduction of the standard \(\ell _2-\)variational formulation of the largest eigenvalue via the classical Motzkin-Straus theorem [37], which could be of independent interest.
Similar content being viewed by others
Notes
\({\mathbb {N}}\) will be used to denote the set of natural numbers, and \({\mathbb {N}}_{\ge k}\) to denote all the natural numbers bigger equal to k.
Throughout the paper, o(1) will be used to denote functions of n that tend to 0 as n tends to infinity. However we will also need to deal with quantities that go to zero as \(\delta \) converges to infinity, which would be denoted by \(o_\delta (1).\)
References
Alt, J., Ducatez, R., Knowles, A.: Extremal eigenvalues of critical erdős-rényi graphs. arXiv preprint arXiv:1905.03243, (2019)
Ben Arous, G., Dembo, A., Guionnet, A.: Aging of spherical spin glasses. Probab. Theory Relat. Fields 120(1), 1–67 (2001)
Ben Arous, G., Guionnet, A.: Large deviations for Wigner’s law and Voiculescu’s non-commutative entropy. Probab. Theory Relat. Fields 108(4), 517–542 (1997)
Augeri, Fanny: Large deviations principle for the largest eigenvalue of wigner matrices without gaussian tails. Electron. J. Probab. 21, 49 (2016)
Augeri, F.: Nonlinear large deviation bounds with applications to traces of wigner matrices and cycles counts in Erdős-Rényi graphs. Ann. Probab. (2020)
Augeri, F., Guionnet, A., Husson, J.: Large deviations for the largest eigenvalue of sub-gaussian matrices. arXiv preprint arXiv:1911.10591 (2019)
Austin, T.: The structure of low-complexity gibbs measures on product spaces. Ann. Probab. 47(6), 4002–4023 (2019)
Bandeira, A.S., Van Handel, R.: Sharp nonasymptotic bounds on the norm of random matrices with independent entries. Ann. Probab. 44(4), 2479–2506 (2016)
Basak, A., Basu, R.: Upper tail large deviations of the cycle counts in Erdős-Rényi graphs in the full localized regime. arXiv:1912.11410 (2019)
Basak, A., Mukherjee, S.: Universality of the mean-field for the potts model. Probab. Theory Relat. Fields 168(3–4), 557–600 (2017)
Benaych-Georges, F., Bordenave, C., Knowles, A.: Largest eigenvalues of sparse inhomogeneous erdős-rényi graphs. Ann. Probab. 47(3), 1653–1676 (2019)
Benaych-Georges, F., Bordenave, C., Knowles, A.: Spectral radii of sparse random matrices. In Annales de l’Institut Henri Poincaré, Probabilités et Statistiques, vol. 56, pp. 2141–2161. Institut Henri Poincaré (2020)
Bhattacharya, B.B, Bhattacharya, S., Ganguly, S.: Spectral edge in sparse random graphs: Upper and lower tail large deviations. Ann. Probab. 49(4), 1847–1885 (2021)
Bhattacharya, B.B., Ganguly, S.: Upper tails for edge eigenvalues of random graphs. SIAM J. Discret. Math. (to appear) (2020)
Bhattacharya, B.B., Ganguly, S., Lubetzky, E., Zhao, Y.: Upper tails and independence polynomials in random graphs. Adv. Math. 319, 313–347 (2017)
Bollobás, B.: Random Graphs, vol. 73. Cambridge University Press, Cambridge (2001)
Bordenave, C., Caputo, P.: A large deviation principle for Wigner matrices without gaussian tails. Ann. Probab. 42(6), 2454–2496 (2014)
Bordenave, C., Caputo, P.: Large deviations of empirical neighborhood distribution in sparse random graphs. Probab. Theory Relat. Fields 163(1–2), 149–222 (2015)
Bordenave, C., Sen, A., Virág, B.: Mean quantum percolation. J. Eur. Math. Soc. 19(12), 3679–3707 (2017)
Chatterjee, S.: Superconcentration and Related Topics, vol. 15. Springer, New York (2014)
Chatterjee, S., Dembo, A.: Nonlinear large deviations. Adv. Math. 299, 396–450 (2016)
Chatterjee, S., Varadhan, S.R.S.: The large deviation principle for the Erdős-Rényi random graph. Eur. J. Combin. 32(7), 1000–1017 (2011)
Chatterjee, S., Varadhan, S.R.S.: Large deviations for random matrices. Commun. Stoch. Anal. 6(1), 1–13 (2012)
Cook, N., Dembo, A.: Large deviations of subgraph counts for sparse Erdős-Rényi graphs. arXiv:1809.11148 (2018)
Dembo, A., Zeitouni, O.: Large Deviations Techniques and Applications, volume 38 of Stochastic Modelling and Applied Probability. Springer, Berlin (2010). Corrected reprint of the second (1998) edition
Eldan, R.: Gaussian-width gradient complexity, reverse log-Sobolev inequalities and nonlinear large deviations. Geom. Funct. Anal. (to appear) (2018)
Erdős, L., Knowles, A., Yau, H.-T., Yin, J.: Spectral statistics of erdős-rényi graphs II: eigenvalue spacing and the extreme eigenvalues. Commun. Math. Phys. 314(3), 587–640 (2012)
Erdős, L., Knowles, A., Yau, H.-T., Yin, J.: Spectral statistics of erdős-rényi graphs I: local semicircle law. Ann. Probab. 41(3B), 2279–2375 (2013)
Frieze, A., Karoński, M.: Introduction to Random Graphs. Cambridge University Press, Cambridge (2016)
Guionnet, A., Husson, J.: Large deviations for the largest eigenvalue of rademacher matrices. Ann. Probab. (to appear) (2020)
Harel, M., Mousset, F., Samotij, W.: Upper tails via high moments and entropic stability. arXiv:1904.08212 (2019)
Krivelevich, M., Sudakov, B.: The largest eigenvalue of sparse random graphs. Comb. Probab. Comput. 12(1), 61–72 (2003)
Latała, R.: Some estimates of norms of random matrices. Proc. Am. Math. Soc. 133(5), 1273–1282 (2005)
Latała, R., van Handel, R., Youssef, P.: The dimension-free structure of nonhomogeneous random matrices. Invent. Math. 214(3), 1031–1080 (2018)
Lubetzky, E., Zhao, Y.: On replica symmetry of large deviations in random graphs. Rand. Struct. Algorith. 47(1), 109–146 (2015)
Lubetzky, E., Zhao, Y.: On the variational problem for upper tails in sparse random graphs. Rand. Struct. Algorith. 50(3), 420–436 (2017)
Motzkin, T.S., Straus, E.G.: Maxima for graphs and a new proof of a theorem of Turán. Can. J. Math. 17, 533–540 (1965)
Reimer, D.: Proof of the van den berg-kesten conjecture. Comb. Probab. Comput. 9(1), 27–32 (2000)
Seginer, Y.: The expected norm of random matrices. Comb. Probab. Comput. 9(2), 149–166 (2000)
Tikhomirov, K., Youssef, P.: Outliers in spectrum of sparse wigner matrices. Rand. Struct. Algorith. (2020)
Van Handel, R.: On the spectral norm of gaussian random matrices. Trans. Am. Math. Soc. 369(11), 8161–8178 (2017)
Yan, J.: Nonlinear large deviations: Beyond the hypercube. Ann. Appl. Probab. (to appear) (2020)
Acknowledgements
The authors thank an anonymous referee for several useful comments that helped improve the article. They also thank Noga Alon for pointing out the classical reference [37]. SG is partially supported by NSF grant DMS-1855688, NSF CAREER Award DMS-1945172 and a Sloan Research Fellowship. KN is supported by the National Research Foundation of Korea (NRF-2019R1A6A1A10073887, NRF-2019R1A5A1028324). This work was initiated when SG was participating in the Probability, Geometry, and Computation in High Dimensions program at the Simons Institute in Fall 2020.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix A. Key estimates
Appendix A. Key estimates
In this appendix, we include the outstanding proofs of basic properties about Gaussian random variables. as well as the proof of Lemma 4.2 involving a straightforward application of Chernoff’s bound.
Proof of Lemma 4.1
Recalling the basic tail bounds from (24), for some constant \(c_1>0\),
Similarly, for some constant \(c_2>0\),
\(\square \)
Proof of Lemma 4.2
We use the Chernoff’s bound for Bernoulli variables for \(q>p\):
where \(I_p(x): = x\log \frac{x}{p}+(1-x) \log \frac{1-x}{1-p}\) is the relative entropy function. Thus,
Using \(\log (1+x) \ge \frac{x}{2}\) for small positive x,
Hence, by (217) and (218), there exists a constant \(c>0\) such that for sufficiently large n,
This implies that
which concludes the proof. \(\square \)
Proof of Lemma 5.1
Recall that we are aiming to show
and in particular, for any \(a,b,c>0\), if \( m \le b\frac{\log n}{\log \log n}+c\) and \(L = a\log n\), then, for any \(\gamma >0\), for sufficiently large n,
By exponential Chebyshev’s bound, for any \(t>0\),
Using the lower bound for the tail (24), the probability density function of \({\tilde{Y}}\), denoted by \({\tilde{f}}(x)\) for \(|x| \ge \sqrt{\varepsilon \log \log n} \), satisfies
Hence, using the upper bound for the tail (24), by making a change of variable \(x = \frac{1}{\sqrt{1-2t}}y\),
Applying this to (220),
We take \(t = \frac{1}{2} (1- \frac{m}{ L } ) < \frac{1}{2}\) (recall that \(L>m\)) in order to balance two terms \(e^{-tL}\) and \( \frac{1}{(1-2t)^m} \). We conclude the proof of (38).
We now show (219). We first check that for any \(L>0\), a function \(x\mapsto (\frac{L}{x})^x\) is increasing on \((0, \frac{L}{e})\). This is because the derivative of \(x \log (\frac{L}{x})\), which is given by \(\log (\frac{L}{x}) - 1\), is positive for \(x\in (0, \frac{L}{e})\). Hence, for any \(\gamma >0\), for sufficiently large n, the LHS of (219) is bounded by
Here, we used the fact that for large n, \( (c_1 \log \log n)^{c_2 \frac{\log n}{\log \log n}} \le n^{\frac{\gamma }{2}}. \)\(\square \)
Rights and permissions
Springer Nature or its licensor holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Ganguly, S., Nam, K. Large deviations for the largest eigenvalue of Gaussian networks with constant average degree. Probab. Theory Relat. Fields 184, 613–679 (2022). https://doi.org/10.1007/s00440-022-01164-7
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00440-022-01164-7