OctoInformBeta

OpenAI preprint, September 2026 · §4. A uniform estimate for cell moments

Lemma 4.2: Uniform marginals and a density bound determined by the geometric mean of the body volumes

Couplings controlling the midpoint law of convex bodies.

Hover over a blue or green term in the text, or a node in the graph, to find it in the other. Select it to read its declaration both in Lean and in words.

The Setting

Two uniformly distributed points in convex bodies can be correlated through a coupling. The quantity of interest is the law of their midpoint relative to Lebesgue measure.

Let E=ℝdE=\mathbb{R}^d, where dd is a nonnegative integer, and let λ\lambda be Lebesgue measure on EE. Write |A|=λ(A)|A|=\lambda(A) for the volume of a set A⊂EA\subset E. The uniform measure ρA\rho_A associated with AA is the scaled restriction of Lebesgue measure given by (1): ρA=|A|−1λ↾A.\rho_A=|A|^{-1}\lambda\!\restriction_A.(1)✦ For the two compact convex sets KK and K′K', write ρ=ρK\rho=\rho_K and ρ′=ρK′\rho'=\rho_{K'}.

A coupling π\pi of probability measures μ\mu and ν\nu on EE is a probability measure on E×EE\times E whose coordinate marginals are μ\mu and ν\nu, as in (2), where pr⁡1\operatorname{pr}_1 and pr⁡2\operatorname{pr}_2 are the coordinate projections: (pr⁡1)#π=μ,(pr⁡2)#π=ν.(\operatorname{pr}_1)_{\#}\pi=\mu, \qquad (\operatorname{pr}_2)_{\#}\pi=\nu.(2)✦ The midpoint law νπ\nu_\pi of a coupling π\pi is its pushforward by the midpoint map m:E×E→Em:E\times E\to E, defined in (3): m(x,y)=x+y2,νπ=m#π.m(x,y)=\frac{x+y}{2}, \qquad \nu_\pi=m_{\#}\pi.(3)

The Claim

Theorem 1. Let K,K′⊂EK,K'\subset E be compact convex sets with nonempty interiors. There is a coupling π\pi of ρ\rho and ρ′\rho' satisfying the measure bound (4): |K||K′|νπ≤λ.\sqrt{|K|\,|K'|}\,\nu_\pi\leq\lambda.(4)✦

Theorem 2. Let K,K′⊂EK,K'\subset E be compact convex sets of positive Lebesgue volume. There is a coupling π\pi of ρ\rho and ρ′\rho' satisfying the measure bound (5): |K||K′|νπ≤λ.\sqrt{|K|\,|K'|}\,\nu_\pi\leq\lambda.(5)✦

What Is Checked

The statements concern normalized restrictions of Lebesgue measure to compact convex sets in finite-dimensional real Euclidean space, with nonempty interiors or positive volume, and probability couplings with the prescribed coordinate marginals.