OctoInformBeta

OpenAI preprint, September 2026 · §1. Introduction

§1 (setting): one-homogeneity, nonzero solutions and flatness: Definitions under positive dilations and almost-everywhere equality

Homogeneity and flatness of Bernoulli profiles.

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

For a one-phase Bernoulli profile, scaling about the origin and comparison with half-space solutions describe its homogeneity and flatness. Fix a dimension d∈ℕd\in\mathbb{N} and write ℝd\mathbb{R}^d for real Euclidean space, with norm |⋅||\cdot| and inner product x⋅ex\cdot e.✦ Almost-everywhere statements refer to Lebesgue measure on ℝd\mathbb{R}^d.

The Definition

Definition 1. A one-homogeneous function u:ℝd→ℝu:\mathbb{R}^d\to\mathbb{R} is a function satisfying the scaling condition (1): for every r>0,u(rx)=ru(x)for almost every x∈ℝd.\text{for every }r>0,\qquad u(rx)=r\,u(x)\quad\text{for almost every }x\in\mathbb{R}^d.(1)✦

A nonzero function u:ℝd→ℝu:\mathbb{R}^d\to\mathbb{R} is a function that is not almost everywhere zero, as specified by (2): ¬(u(x)=0 for almost every x∈ℝd).\neg\bigl(u(x)=0\text{ for almost every }x\in\mathbb{R}^d\bigr).(2)✦

A flat function u:ℝd→ℝu:\mathbb{R}^d\to\mathbb{R} is a function agreeing almost everywhere with a half-space profile in a unit direction, as in (3): there exists e∈ℝd with |e|=1 such thatu(x)=(x⋅e)+=max⁡{x⋅e,0}for almost every x∈ℝd.\begin{gathered} \text{there exists }e\in\mathbb{R}^d\text{ with }|e|=1\text{ such that}\\ u(x)=(x\cdot e)_{+}=\max\{x\cdot e,0\} \quad\text{for almost every }x\in\mathbb{R}^d. \end{gathered}(3)✦

What Is Checked

The definitions (1), (2) and (3) are stated for arbitrary functions u:ℝd→ℝu:\mathbb{R}^d\to\mathbb{R} in every dimension d∈ℕd\in\mathbb{N}.