Abstract:We study the regularity of the law of a quadratic form $Q(X,X)$, evaluated in a sequence $X = (X_{i})$ of independent and identically distributed random variables, when $X_{1}$ can be expressed as a sufficiently smooth function of a Gaussian field. This setting encompasses a large class of important and frequently used distributions, such as, among others, Gaussian, Beta, for instance uniform, Gamma distributions, or else any polynomial transform of them. Let us present an emblematic application. Take $X = (X_{i})$ a sequence of independent and identically distributed centered random variables, with unit variance, following such distribution. Consider also $(Q_{n})$ a sequence of quadratic forms, with associated symmetric Hilbert--Schmidt operators $(\mathsf{A}^{(n)})$. Assume that $\operatorname{Tr}[ (\mathsf{A}^{(n)})^{2} ] = 1/2$, $\mathsf{A}^{(n)}_{ii} =0$, and the spectral radius of $\mathsf{A}^{(n)}$ tends to $0$. Then, $(Q_{n}(X))$ converges in a strong sense to the standard Gaussian distribution. Namely, all derivatives of the densities, which are well-defined for $n$ sufficiently large, converge uniformly on $\mathbb{R}$ to the corresponding derivatives of the standard Gaussian density. While classical methods, from Malliavin calculus or $\Gamma$-calculus, generally consist in bounding negative moments of the so-called \emph{carré du champ} operator $\Gamma(Q(X),Q(X))$, we provide a new paradigm through a second-order criterion involving the eigenvalues of a Hessian-type matrix related to $Q(X)$. This Hessian is built by iterating twice a tailor-made gradient, the \emph{sharp operator} $\sharp$, obtained via a Gaussian representation of the carré du champ. We believe that this method, recently developed by the authors in the current paper and in their companion paper [AoP 52 n°3 (2024)] , is of independent interest and could prove useful in other settings.

A central limit theorem for generalized quadratic forms

Joint CLT for Several Random Sesquilinear Forms with Applications to Large-Dimensional Spiked Population Models

Central Limit Theorem of Random Quadratics Forms Involving Random Matrices

Central limits from generating functions

A Limit Result for Self-Normalized Random Sums

An almost everywhere central limit theorem

Central Limit Theorems for Weighted Quadratic Forms of Dependent Processes with Applications in Specification Testing

Lindeberg's Central Limit Theorems for Martingale-Like Sequences under Sub-Linear Expectations

Regularity of laws via Dirichlet forms: application to quadratic forms in independent and identically distributed random variables

Regularity of laws via Dirichlet forms -- Application to quadratic forms in independent and identically distributed random variables

On strong versions of the central limit theorem

Fluctuations of Quadratic Chaos

Central Limit Theorem for Linear Eigenvalue Statistics of non-Hermitian Random Matrices

Some Notes on Quantitative Generalized CLTs with Self-Decomposable Limiting Laws by Spectral Methods

A central limit theorem for coefficients of L-functions in short intervals

Norm Convergence Rate for Multivariate Quadratic Polynomials of Wigner Matrices

Limit theorems for a strongly irreducible product of independent random matrices under optimal moment assumptions

Universality for Random Matrices

Multi-Point Functional Central Limit Theorem for Wigner Matrices

A central limit theorem associated with a sequence of positive line bundles

Central Limit Theorems for Exchangeable Random Variables When Limits Are Scale Mixtures of Normals