betaFeedback

WorksMathematics

Sharp Gaussian Brunn-Minkowski inequalities

Abstract

We prove Gaussian Brunn-Minkowski inequalities for multiple sets. These are of the type: ∏i=1IγΣi(Ki)λ~i≤CsharpγΓ(∑i=1IλiKi),\prod_{i=1}^I γ_{Σ_i}(K_i)^{\tilde λ_i} \le C_{\text{sharp}}γ_Γ\Big(\sum_{i=1}^I λ_i K_i\Big), with convex sets Ki⊂RnK_i \subset \mathbb{R}^n, whose barycenter bKi,Σib_{K_i,Σ_i} with respect to the Gaussian measure γΣiγ_{Σ_i} is zero, where λi,λ~i∈(0;1)λ_i, \tilde λ_i \in (0;1) such that ∑i=1Iλi=∑i=1Iλ~i=1.\sum_{i=1}^I λ_i=\sum_{i=1}^I \tilde λ_i=1. In addition, we provide natural conditions on the matrices Γ,Σ1,⋯ ,ΣI∈Sym>0(n)Γ, Σ_1,\cdots,Σ_I \in \mathrm{Sym}_{>0}(n), which are given through a pointwise inequality involving the Karcher mean. Furthermore, we study equality cases of the corresponding functional forms in such situations. We also investigate the role of centering in the (B)-property by giving a counterexample for convex bodies with zero Lebesgue barycenters in dimensions at least three and thereby answering negatively a question of Faifman-Polterovich in \cite{Faifman_Polt}. Finally, a new slicing approach is introduced, which allows us to pass from centered convex sets to sets with arbitrary barycenters such that their weighted sum is convex. Based on this idea, we show a new Gaussian Brunn-Minkowski inequality e∣∑i=1IλibKi,In∣2/2nγn(∑i=1IλiKi)1/n≥∑i=1Iλie∣bKi,In∣2/2nγn(Ki)1/n,e^{|\sum_{i=1}^I λ_i b_{K_i,I_n}|^2/2n} γ_n \Big(\sum_{i=1}^I λ_i K_i \Big)^{1/n} \ge \sum_{i=1}^I λ_i e^{|b_{K_i,I_n}|^2/2n} γ_n(K_i)^{1/n}, and consider a further application in the direction of the Gaussian correlation inequality, together with the rigidity for both inequalities.

2020 Mathematics Subject Classification. 52A40 (primary), 60E15, 60G15 (secondary).

Introduction

The celebrated Brunn-Minkowski inequality states that for all K,L⊂RnK,L\subset\mathbb{R}^n measurable sets with positive measure,

∣K+L∣1/n≥∣K∣1/n+∣L∣1/n.|K+L|^{1/n}\ge|K|^{1/n}+|L|^{1/n}.

By homogeneity of the Lebesgue measure, it can be written as

∣(1−t)K+tL∣≥∣K∣1−t∣L∣t∀t∈[0;1].|(1-t)K+tL|\ge|K|^{1-t}|L|^t \quad\forall t\in[0;1].

Its well-known functional form is the Prékopa-Leindler inequality, which says that

∫Rnsup⁡z=tx+(1−t)yf(x)tg(y)1−t dz≥(∫Rnf(x) dx)t(∫Rng(y) dy)1−t,\int_{\mathbb{R}^n}\sup_{z=tx+(1-t)y}f(x)^t g(y)^{1-t}\,dz\ge\left(\int_{\mathbb{R}^n}f(x)\,dx\right)^t\left(\int_{\mathbb{R}^n}g(y)\,dy\right)^{1-t},

for any f,g:Rn→[0;+∞)f,g:\mathbb{R}^n\to[0;+\infty) with finite and positive integrals. The previous functional inequality is known thanks to the pioneering work of Barthe [6] to be a reverse Brascamp-Lieb inequality. We refer the reader to the beautiful survey of [20] and the references therein.

Brascamp-Lieb inequalities were first studied in [10] and are functional inequalities whose sharp constant is completely determined by centered Gaussian functions (commonly known as the Gaussian saturation phenomenon) as proven by Lieb in [29]. These inequalities encompass many important inequalities such as Hölder, Young convolution or Loomis-Whitney and immediately found important applications in the field of convex geometry thanks to the works of Keith Ball (see [4, 5]).

Two natural and closely related questions with regard to such functional or geometric inequalities can then be addressed:

  1. By imposing natural assumptions on the sets or functions considered, could certain Brunn-Minkowski type inequalities become true?

(ii) Can we replace the Lebesgue measure by other kinds of measures for which the statement would still hold, up to the aforementioned assumptions?

These two questions turn out to be highly general, very challenging and have motivated a major direction of research in the field of convex geometry. A first natural assumption on sets (resp. functions) is convexity (resp. log-concavity) but symmetry with respect to the origin (K=−KK=-K for a set and requiring evenness for functions) proves to be also a crucial assumption.

The usual measures considered for the validity of such inequalities are log-concave measures and more specifically, Gaussian measures. Recall that for Σ∈Sym⁡>0(n)\Sigma\in\operatorname{Sym}_{>0}(n), we write γΣ\gamma_{\Sigma} for the centered Gaussian probability measure with covariance Σ\Sigma and

γΣ:=1det⁡(2πΣ)e−12⟨Σ−1x,x⟩ dx.\gamma_{\Sigma}:=\frac{1}{\sqrt{\det(2\pi\Sigma)}}e^{-\frac{1}{2}\langle\Sigma^{-1}x,x\rangle}\,dx.

A first example of such inequality is linked to the (B)-theorem, a problem proposed by Banaszczyk and popularized by Łatala in [26], then subsequently solved by Cordero-Erausquin, Fradelizi and Maurey in [12], where it is proven that if K⊂RnK\subset\mathbb{R}^{n} is a symmetric convex set, then the function

(t1,⋯ ,tn)↦γn(ediag⁡(t1,⋯ ,tn)K)(t_{1},\cdots,t_{n})\mapsto\gamma_{n}(e^{\operatorname{diag}(t_{1},\cdots,t_{n})}K)

is log-concave on Rn\mathbb{R}^{n}.

The proof relies on Poincaré-type inequalities for γK\gamma_{K}, the normalized restriction of γn\gamma_{n} to the symmetric convex set KK, and the so-called Hörmander’s L2L^{2} method.

This method has become crucial to study Brunn-Minkowski type inequalities for the Gaussian measure and to understand (B)-properties of log-concave measures. First, by the Prékopa-Leindler inequality, for all K,L⊂RnK,L\subset\mathbb{R}^{n} with positive Gaussian measure,

γn((1−t)K+tL)≥γn(K)1−tγn(L)t∀t∈[0;1],\gamma_{n}((1-t)K+tL)\geq\gamma_{n}(K)^{1-t}\gamma_{n}(L)^{t}\quad\forall t\in[0;1],

however, as homogeneity fails for the Gaussian measure, this does not recover an additive-type inequality of the form

γn((1−t)K+tL)1/n≥(1−t)γn(K)1/n+tγn(L)1/n∀t∈[0;1].(1)\gamma_{n}((1-t)K+tL)^{1/n}\geq(1-t)\gamma_{n}(K)^{1/n}+t\gamma_{n}(L)^{1/n}\quad\forall t\in[0;1]. \tag*{(1)}

In fact, such a statement was conjectured by Gardner and Zvavitch in [21] for convex sets K,LK,L containing the origin but a counterexample was provided in [35]. Thanks to the works of Kolesnikov-Livshyts [24] and later of Eskenazis-Moschidis [18], (1.1) was proven for symmetric convex sets and later, Cordero-Erausquin and Rotem in [15] showed that such a Brunn-Minkowski inequality holds for all rotationally invariant log-concave measures. Additionally, they proved that the (B)-property, i.e. for every symmetric convex K⊂RnK\subset\mathbb{R}^{n} and every symmetric matrix AA,

t↦μ(etAK)t\mapsto\mu(e^{tA}K)

is log-concave

holds for any rotationally invariant log-concave measure μ\mu.

For more recent works using such approaches, we refer the reader to [13, 14]. Let us note that in the case of the standard Gaussian measure, a proof of the (B)-property (or in other words, the strong form of the Gaussian (B) inequality) was already fully shown by Brascamp and Lieb in [11] and employs the rotation trick, a crucial argument in the proof of the Gaussian saturation, together with the variance Brascamp-Lieb inequality applied to a linear odd function (see the appendix of [13]).

In fact, only focusing on the (B)-theorem for Gaussian measures with arbitrary covariances is linked to the fundamental log Brunn-Minkowski problem, first formulated by Böröczky, Lutwak, Yang and Zhang in [8], connected to important problems of uniqueness of the cone measures and Monge-Ampère, see [9] and recently announced to be solved by OpenAI in [36]. By [38], the log Brunn-Minkowski inequality, which is yet an example of pivotal improvement compared to the Brunn-Minkowski inequality under convexity and evenness,

∣⋂u∈Sn−1{x∈Rn:⟨x,u⟩≤hK(u)1−thL(u)t}∣≥∣K∣1−t∣L∣t,\left|\bigcap_{u\in\mathbb{S}^{n-1}}\left\{x\in\mathbb{R}^{n}:\langle x,u\rangle\leq h_{K}(u)^{1-t}h_{L}(u)^{t}\right\}\right|\geq|K|^{1-t}|L|^{t},

where hK(x):=sup⁡y∈K⟨x,y⟩h_{K}(x):=\sup_{y\in K}\langle x,y\rangle, is equivalent to the fact that for all Σ∈Sym⁡>0(n)\Sigma\in\operatorname{Sym}_{>0}(n) and for every diagonal matrix A∈Sym⁡(n)A\in\operatorname{Sym}(n),

t↦γΣ(etA[−1;1]n)t\mapsto\gamma_{\Sigma}(e^{tA}[-1;1]^{n})

is log-concave.

However, in [16], it is shown that t↦γΣ(etAK)t\mapsto\gamma_{\Sigma}(e^{tA}K) may not be necessarily log-concave for a symmetric convex body K⊂RnK\subset\mathbb{R}^{n} and a diagonal matrix AA.

Lastly, another recent way of proving (1) has appeared in [1] using information-theoretic tools and constructing a transport coupling, which satisfies the entropic inequality corresponding to the Gaussian Brunn-Minkowski inequality. Building on [2], Han-Zhu in [23] proved that the displacement interpolation between two 1-strongly log-concave probability measures has also Poincaré constant at most 1, in particular settling [2, Question 1.6], and deduced a Gaussian Brunn-Minkowski inequality for two sets: for K1,K2⊂RnK_1,K_2 \subset\mathbb{R}^n convex sets with γn(K1),γn(K2)>0\gamma_n(K_1),\gamma_n(K_2)>0,

e∣(1−t)bK1+tbK2∣2/2n⋅γn((1−t)K1+tK2)1/n≥(1−t)e∣bK1∣2/2nγn(K1)1/n+te∣bK2∣2/2nγn(K2)1/n,(2)e^{\lvert(1-t)b_{K_1}+tb_{K_2}\rvert^2/2n}\cdot\gamma_n\left((1-t)K_1+tK_2\right)^{1/n}\ge(1-t)e^{\lvert b_{K_1}\rvert^2/2n}\gamma_n(K_1)^{1/n}+te^{\lvert b_{K_2}\rvert^2/2n}\gamma_n(K_2)^{1/n}, \tag*{(2)}

where

bKi:=1γn(Ki)∫Kix dγn(x) for i=1,2.b_{K_i}:=\frac{1}{\gamma_n(K_i)}\int_{K_i}x\,\mathrm{d}\gamma_n(x)\text{ for }i=1,2.

Let us now come back to the relationship with Brascamp-Lieb inequalities. As explained previously, these functional inequalities have important consequences but as they hold for any functions, it is again worth considering whether by restricting functions to for example log-concave and even functions, one could have interesting inequalities. Nakamura and Tsuji in [33] introduced such a viewpoint of improvement under evenness and log-concavity in the theory of Brascamp-Lieb inequalities, first through inverse Brascamp-Lieb inequalities considered in the general case by [7]. Later, Milman in [30] realized that such a formulation yielded the celebrated Gaussian correlation inequality (see [37] [27] for the original arguments) and Nakamura and Tsuji in [34] pushed further this idea to obtain a generalized Gaussian correlation inequality with multiple centered convex sets. These works culminated in [32] to obtain a general Gaussian Forward-Reverse Brascamp-Lieb (FRBL) with the required assumptions of log-concavity and centering (the Lebesgue case was known from [17]), which allowed to prove the strong Gaussian correlation inequality.

Our main contributions in this paper are as follows: we first determine the sharp constants in Gaussian Brunn–Minkowski inequalities for arbitrary finite families of convex sets centered with respect to possibly different Gaussian measures. We characterize equality under Karcher mean conditions and in the associated functional interpolation inequalities. We also introduce a slicing method that yields Gaussian inequalities with barycenter corrections, including Brunn-Minkowski inequalities for measurable sets whose weighted Minkowski sum is convex and correlation inequalities for convex sets with a common Gaussian barycenter. Finally, we investigate the role of centering in the (B)-property, constructing counterexamples and showing that optimization over translations restores log-concavity.

We would like to note that Theorem 1.1 was obtained independently of [31]. In the two-input case, Theorem 1.1 coincides with the sharp heterovariant Gaussian Brunn-Minkowski inequality established by Milman, Nakamura and Tsuji in the second version of [31]. Both works employ the Gaussian Forward-Reverse Brascamp-Lieb theorem of Milman, Nakamura and Tsuji [32] to reduce the sharp-constant problem to a matrix optimization one. Here, the optimization is subsequently treated for arbitrary finite families using Schur complements and induction on the dimension. The results below also provide functional inequalities and equality characterizations under the Karcher mean conditions.

Gaussian inequalities, convexity and centering

Let I≥2I\ge2 and λ=(λ1,…,λI)∈ΔI∘\lambda=(\lambda_1,\ldots,\lambda_I)\in\Delta_I^\circ, where the relative interior of the unit simplex is denoted by

ΔI∘={(λ1,…,λI):λi>0, ∑i=1Iλi=1}.\Delta_I^\circ=\left\{(\lambda_1,\ldots,\lambda_I):\lambda_i>0,\ \sum_{i=1}^{I}\lambda_i=1\right\}.

If ai>0a_i>0 for 1≤i≤I1\le i\le I, denote their weighted power mean by

Mpλ(a1,…,aI)={(∑i=1Iλiaip)1/p,0<p<+∞,∏i=1Iaiλi,p=0,min⁡1≤i≤Iai,p=+∞.\mathcal{M}_p^\lambda(a_1,\ldots,a_I)= \begin{cases} \left(\sum_{i=1}^{I}\lambda_i a_i^p\right)^{1/p}, & 0<p<+\infty,\\ \prod_{i=1}^{I}a_i^{\lambda_i}, & p=0,\\ \min_{1\le i\le I}a_i, & p=+\infty. \end{cases}

for any p∈[0;+∞]p \in[0;+\infty] and λ∈ΔI∘\lambda\in\Delta_I^\circ. Moreover, the mean is set to zero if any aia_i is zero.

We will write for a given matrix Γ∈Sym⁡(Rn)\Gamma\in\operatorname{Sym}(\mathbb{R}^n),

ΓE=PEΓ∣E,\Gamma_E=P_E\Gamma|_E,

where PEP_E denotes the orthogonal projection onto a subspace E⊂RnE \subset\mathbb{R}^n.

Our first main result is as follows.

References

  1. [28]Lawson Jimmie and Lim Yongdo, ‘Karcher means and Karcher equations of positive operators’, in: Trans. Amer. Math. Soc. Ser. B 1 (2014), pp. 1–22, DOI: 10.1090/S2330-0000-2014-00003-4.DOI
  2. [29]Lieb Elliott H., ‘Gaussian kernels have only Gaussian maximizers’, in: Invent. Math. 102.1 (1990), pp. 179–208, DOI: 10.1007/BF01233426.DOI
  3. [30]Milman Emanuel, ‘Gaussian correlation via Inverse Brascamp-Lieb’, in: Probability Theory and Related fields (2025), DOI: 10.1007/s00440-025-01445-x.DOI
  4. [31]Milman Emanuel, Nakamura Shohei and Tsuji Hiroshi, Gaussian inequalities with different covariance matrices - the Heterovariant Gaussian Brunn-Minkowski inequality, 2026, eprint: 2610.13596.arxiv.org/abs/2610.13596
  5. [32]Milman Emanuel, Nakamura Shohei and Tsuji Hiroshi, The Gaussian Conjugate Rogers-Shephard Inequality, 2026, eprint: 2602.07981.arxiv.org/abs/2602.07981
  6. [33]Nakamura Shohei and Tsuji Hiroshi, ‘A generalized Legendre duality relation and Gaussian saturation’, in: Invent. Math. 243.2 (2026), pp. 607–655, DOI: 10.1007/s00222-025-01382-5.DOI
  7. [34]Nakamura Shohei and Tsuji Hiroshi, The Gaussian correlation inequality for centered convex sets and the case of equality, 2025, eprint: 2504.04337.arxiv.org/abs/2504.04337
  8. [35]Nayar Piotr and Tkocz Tomasz, ‘A note on a Brunn-Minkowski inequality for the Gaussian measure’, in: Proc. Amer. Math. Soc. 141.11 (2013), pp. 4027–4030, DOI: 10.1090/S0002-9939-2013-11609-6.DOI
  9. [36]OpenAI, The logarithmic Brunn-Minkowski conjecture, https://github.com/openai/math/blob/main/preprints/The-logarithmic-Brunn-Minkowski-conjecture-September-23-2026/paper.pdf, 2026.
  10. [37]Rోయen Thomas, ‘A simple proof of the Gaussian correlation conjecture extended to some multivariate gamma distributions’, in: Far East J. Theor. Stat. 48.2 (2014), pp. 139–145.arxiv.org/abs/1408.1028
  11. [38]Saroglou Christos, ‘More on logarithmic sums of convex bodies’, in: Mathematika 62.3 (2016), pp. 818–841, DOI: 10.1112/S0025579316000061.DOI

Paper details

vlnc:W_k9wkw9u7
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

a commercial AI tool

View statement
Looking for related papers…
Contents