CLT for β-ensembles with Freud weights, application to the KLS conjecture in Schatten balls

Basic Information

• Paper ID: 2511.05386
• Title: CLT for β-ensembles with Freud weights, application to the KLS conjecture in Schatten balls
• Authors: Charlie Dworaczek Guera, Ronan Memin, Michel Pain
• Classification: math.PR (Probability Theory), math.FA (Functional Analysis), math.MG (Metric Geometry)
• Publication Date: November 10, 2025
• Paper Link: https://arxiv.org/abs/2511.05386

Abstract

This paper investigates β-ensembles (or one-dimensional log gases) with Freud weights, namely systems with potential function $|x|^p$ ($p \geq 2$). When $p \in (2,3]$, the potential function does not belong to the $\mathcal{C}^3$ class, rendering most existing results in the literature inapplicable. In this singular setting, the authors establish a central limit theorem for linear statistics with general test functions and compute the subleading correction term of the free energy. The research strategy is based on establishing an optimal local law. These results enable the authors to provide a large-$N$ expansion of the logarithmic volume of the unit ball of $N \times N$ self-adjoint matrices in the $p$-Schatten norm and perform consistency checks on the KLS conjecture.

Research Background and Motivation

Research Problems

This paper addresses two core problems:

1. Central Limit Theorem for Singular Potentials: For β-ensembles with Freud weights $V(x) = c_p|x|^p$ ($p \geq 2$), when $p \in (2,3]$ the potential function lacks $\mathcal{C}^3$ regularity at the origin, causing traditional methods to fail.
2. Verification of the KLS Conjecture: Utilizing β-ensemble theory to verify the Kannan-Lovász-Simonovits conjecture in the special case of Schatten balls.

Importance of the Problems

• Random Matrix Theory: β-ensembles are core models in random matrix theory, corresponding to the eigenvalue distributions of classical ensembles such as GOE ($\beta=1$), GUE ($\beta=2$), and GSE ($\beta=4$).
• Convex Geometry Applications: Results directly apply to important conjectures in high-dimensional convex geometry, particularly the KLS conjecture, which has been an important open problem for 30 years in the field.
• Quantum Field Theory Connection: The study of microscopic behavior near critical points in quantum chromodynamics requires theory for such singular potentials.

Limitations of Existing Methods

1. Regularity Requirements: Existing CLT results typically require potential functions $V \in \mathcal{C}^6$ and test functions $f \in \mathcal{C}^4$.
2. Technical Obstacles: When $p \in (2,3]$, the density of the equilibrium measure diverges near the endpoints, causing the inverse of the principal operator $\Psi_\alpha = \Xi_\alpha^{-1}[f]$ to have only $\mathcal{C}^2$ regularity.
3. Method Failure: The traditional principal operator method requires $\|\Psi''_\alpha\|_\infty < \infty$, but this condition fails in the singular case.

Core Contributions

1. CLT for Singular Potentials: First establishment of a central limit theorem with general test functions in the singular case $p \in (2,3]$.
2. Optimal Local Law: Establishment of an optimal local law of the form $\mathbb{E}[|s_N(z) - s_{V_\alpha}(z)|^q] \leq (Cq^{64})^q/(Ny)^q$.
3. Subleading Term of Free Energy: Computation of the precise asymptotic expansion of the partition function $\log Z_N$ up to the $o(N)$ term.
4. Schatten Ball Volume Formula: Provision of a complete asymptotic expansion of the logarithmic volume of the $p$-Schatten unit ball.
5. KLS Conjecture Verification: Proof that the KLS conjecture holds for the function $f(X) = \text{Tr}(X^r)^q$ on Schatten balls.

Methodology Details

Problem Setup

Consider the β-ensemble probability measure: $d\mathbb{P}(\lambda_1,\ldots,\lambda_N) = \frac{1}{Z_N}\prod_{1\leq i<j\leq N}|\lambda_i-\lambda_j|^\beta \cdot e^{-\frac{\beta N}{2}\sum_{i=1}^N V(\lambda_i)}d\lambda_1\cdots d\lambda_N$

where $V(x) = c_p|x|^p$, and the goal is to study the fluctuation properties of the empirical measure $\mu_N = \frac{1}{N}\sum_{i=1}^N\delta_{\lambda_i}$.

Core Technical Framework

1. Interpolation Strategy

Introduction of an interpolated potential function: $V_\alpha(x) = \alpha c_p|x|^p + (1-\alpha)2x^2, \quad \alpha \in [0,1]$

This allows continuous variation from the quadratic potential ($\alpha=0$) to the Freud potential ($\alpha=1$).

2. Loop Equations and Local Law

Innovative Loop Equations: Unlike traditional methods, the authors combine two integral identities to obtain: $\mathbb{E}_\alpha\left[\left(P_\alpha(z) + \frac{L_N(f_{\alpha,z})}{Nz} + \frac{1}{N}\left(\frac{2}{\beta}-1\right)s'_N(z)\right)P_\alpha(z)^{q-1}\overline{P_\alpha(z)^q}\right] + \text{other terms} = 0$

where $P_\alpha(z) = s_N(z)^2 + \frac{g_\alpha(z)}{z}s_N(z) + \frac{h_\alpha(z)}{z}$.

Key Innovation: Replacement of $\frac{V'_\alpha(\lambda)-V'_\alpha(z)}{\lambda-z}$ with $f_{\alpha,z}(\lambda) = \frac{g_\alpha(\lambda)-g_\alpha(z)}{\lambda-z}$, trading the singularity of $1/z$ for additional regularity.

3. Stability Lemma

Proof that when $u$ approximately satisfies the quadratic equation $u^2 + \frac{g_\alpha(z)}{z}u + \frac{h_\alpha(z)}{z} = \zeta$: $|u - s_{V_\alpha}(z)| \wedge |u - \tilde{s}_{V_\alpha}(z)| \leq C\left(\frac{|\zeta|}{|b(z)|} \wedge |\zeta|^{1/2}\right)$

4. Principal Operator Method

The principal operator $\Xi_\alpha$ is defined as: $\Xi_\alpha[\psi](x) = -\frac{1}{2}\psi(x)V'_\alpha(x) + \int_{-1}^1 \frac{\psi(x)-\psi(t)}{x-t}d\mu_{V_\alpha}(t)$

Technical Breakthrough: Through the Helffer-Sjöstrand formula, the regularity requirement is reduced from $\|\psi''\|_\infty < \infty$ to $\|\psi''\|_1 < \infty$.

Technical Innovations

1. Regularization Techniques: Development of regularization methods applicable to singular functions, allowing the Helffer-Sjöstrand formula to remain applicable.
2. Bootstrap Strategy: Progressive improvement of bounds on linear statistics through four-fold iterative application of the local law mechanism.
3. Pseudo-analytic Continuation: Careful complex extension of the function $g(x) = px|x|^{p-2}$ to handle singularities when $p \in (2,3)$.

Main Theorems

Theorem 1.1 (Central Limit Theorem)

Let $p \geq 2$ and $f \in \mathcal{C}^3(\mathbb{R})$ with at most exponential growth. Then: $L_N(f) \xrightarrow{N\to\infty} \mathcal{N}(m_V(f), \sigma^2(f))$

where the mean and variance are given by explicit formulas.

Theorem 1.3 (Free Energy Expansion)

$\frac{1}{N^2\beta}\log Z_N = -\frac{1}{2}\left(\log 2 + \frac{3}{2p}\right) + \frac{\log N}{2N} + \frac{F^{\{-1\}}}{N} + o\left(\frac{1}{N}\right)$

Theorem 1.5 (Optimal Local Law)

There exist $C,\delta_0 > 0$ such that for any $z = x+iy \in \mathcal{R}$: $\mathbb{E}[|s_N(z) - s_V(z)|^q] \leq \frac{(Cq^{64})^q}{(Ny)^q}$

Applications to Schatten Balls

Volume Formula (Corollary 1.9)

$\log|B_E(S_p^N)| = aN^2\log N + bN^2 + cN\log N + dN + o(N)$

where the coefficients $a,b,c,d$ are given by explicit formulas.

KLS Conjecture Verification (Theorem 1.10)

For $f(X) = \text{Tr}(X^r)^q$ (with $r$ even), the KLS conjecture holds on Schatten balls: $\text{Var}_\mu f \leq 4 \cdot \lambda_\mu \cdot \mathbb{E}_\mu[\|\nabla f\|_2^2]$

Experimental Results and Theoretical Verification

As this is a pure mathematical theoretical work, results are verified primarily through rigorous mathematical proofs. Key verifications include:

1. Optimality of Local Law: Proof that the bound $O(1/(Ny))$ is optimal.
2. Interpolation Formula Verification: Verification of the correctness of the free energy formula through $\alpha$-interpolation.
3. Consistency with Known Results: Complete agreement with results from DFGZ23 in the case $p>3$.

Related Work

CLT in Random Matrix Theory

• Johansson (1998): First establishment of CLT for β-ensembles, requiring polynomial potentials.
• Borot-Guionnet (2013): Analytic potential case.
• Bekerman-Leblé-Serfaty (2018): Requiring $V \in \mathcal{C}^6$, $f \in \mathcal{C}^4$.

Local Law Theory

• Bourgade-Mody-Pain (2022): Optimal local law for analytic potentials.
• Bourgade-Erdős-Yau (2014): Suboptimal local law for $V \in \mathcal{C}^4$.

KLS Conjecture

• Kannan-Lovász-Simonovits (1995): Original conjecture.
• Klartag-Lehec (2024,2025): Recent resolution of the variance conjecture and hyperplane conjecture.

Conclusions and Discussion

Main Conclusions

1. Successful extension of β-ensemble CLT theory to the singular potential case.
2. Establishment of new bridges connecting random matrix theory and convex geometry.
3. Provision of new verification cases for the KLS conjecture.

Limitations

1. Regularity Restrictions: Still requires $p \geq 2$; the weakly constrained case $p < 2$ is more difficult.
2. Test Function Requirements: Requires $f \in \mathcal{C}^3$; while weaker than existing results, still restrictive.
3. Limited KLS Verification Scope: Verification only for specific function forms $f(X) = \text{Tr}(X^r)^q$.

Future Directions

1. Extension to $p \in (1,2)$: Study of the weakly constrained regime.
2. Multiple Critical Points: Investigation of more complex potential functions.
3. Complete Resolution of KLS Conjecture: Pursuit of more general verification methods.

In-Depth Evaluation

Strengths

1. Strong Technical Innovation: Clever combination of loop equations, principal operator methods, and the Helffer-Sjöstrand formula.
2. Theoretical Completeness: Complete theoretical chain from local law to CLT to applications.
3. High Application Value: Direct application to important geometric conjectures.
4. Clear Exposition: Rich technical details with clear logical structure.

Weaknesses

1. High Technical Complexity: Requires combination of multiple advanced techniques with high entry barriers.
2. Limited Application Scope: Primarily targeted at specific forms of potentials and test functions.
3. Computational Complexity: Explicit expressions for certain constants are quite involved.

Impact

1. Theoretical Contribution: Fills important gaps in singular potential theory.
2. Methodological Value: Provides new technical frameworks for handling singular problems.
3. Interdisciplinary Impact: Connects probability theory, functional analysis, and convex geometry.

Applicable Scenarios

1. Random Matrix Theory: Extension of existing CLT theory to broader classes of potentials.
2. Convex Geometry: Study of volume and concentration properties of high-dimensional convex bodies.
3. Quantum Physics: Theory of critical phenomena and phase transitions.

References

This paper cites important literature in the field, including:

• Johansson (1998): Foundational work on β-ensemble CLT.
• Bourgade-Mody-Pain (2022): Important reference for optimal local laws.
• Bekerman-Leblé-Serfaty (2018): Key literature on principal operator methods.
• Dadoun-Fradelizi-Guédon-Zitt (2023): Direct predecessor for Schatten ball applications.

This paper demonstrates high technical innovation and complexity, successfully resolving a difficult theoretical problem with important applications in convex geometry. While the technical threshold is high, it makes significant contributions to the development of related fields.