OctoInformBeta

OpenAI preprint, September 2026 · §1. Introduction · §1

Corollary 1.2 [Weak pinned distance conjecture]: For ε>0\varepsilon>0, the largest fraction of pins of an nn-point planar set with fewer than n1−εn^{1-\varepsilon} distinct distances tends to zero

Most pins of a large planar set determine many distances.

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

Take a set PP of nn points in the plane and a point xx of PP, the pin. The other points of PP lie on circles about xx, and the distances the pin determines are the radii of these circles. The question, a weak form of the Erdős pinned distance problem, fixes an exponent ε>0\varepsilon>0 and asks how large a fraction of the points of an nn-point set can be pins determining fewer than n1−εn^{1-\varepsilon} distinct distances, as nn grows.

Distances at a Pin

A configuration PP is a finite set of points in the Euclidean plane ℝ2\mathbb{R}^2, and n=|P|n=|P| is its number of points. The distance set Dx(P)D_x(P) of a configuration PP at a point xx of the plane is the set of distances from xx to the points of PP other than xx: Dx(P)={∥y−x∥:y∈P,y≠x}.D_x(P)=\{\,\|y-x\| : y\in P,\ y\neq x\,\}.(1)✦

A bad pin of a configuration PP at a real exponent ε\varepsilon is a point x∈Px\in P whose distance set has fewer than n1−εn^{1-\varepsilon} elements.✦ For n≥1n\ge1 and real ε\varepsilon, the extremal bad fraction Bn(ε)B_n(\varepsilon) is the supremum, over all configurations PP of nn points, of the fraction of bad pins in PP, and B0(ε)=0B_0(\varepsilon)=0: Bn(ε)=sup|P|=n|{x∈P:|Dx(P)|<n1−ε}|n(n≥1).B_n(\varepsilon)=\sup_{|P|=n}\frac{|\{x\in P : |D_x(P)|<n^{1-\varepsilon}\}|}{n} \qquad(n\ge1).(2)✦

Rich Pairs

The multiplicity kP(x,y)k_P(x,y) of points xx and yy of the plane in a configuration PP is the number of points of PP other than xx on the circle about xx through yy: kP(x,y)=|{z∈P:z≠x,∥z−x∥=∥y−x∥}|.k_P(x,y)=|\{\,z\in P : z\neq x,\ \|z-x\|=\|y-x\|\,\}|.(3)✦

An ss-rich pair of a configuration PP, for a real ss, is an ordered pair (x,y)(x,y) of distinct points of PP with kP(x,y)≥nsk_P(x,y)\ge n^{s}.✦ For n≥2n\ge2 and real ss, the extremal rich-pair fraction Fn(s)F_n(s) is the supremum, over all configurations PP of nn points, of the number of ss-rich pairs of PP divided by n(n−1)n(n-1), and Fn(s)=0F_n(s)=0 for n<2n<2: Fn(s)=sup|P|=n|{(x,y)∈P×P:x≠y,kP(x,y)≥ns}|n(n−1)(n≥2).F_n(s)=\sup_{|P|=n}\frac{|\{(x,y)\in P\times P : x\neq y,\ k_P(x,y)\ge n^{s}\}|}{n(n-1)} \qquad(n\ge2).(4)✦

The Claim

Theorem 1. For every ε>0\varepsilon>0, the extremal bad fraction tends to zero: limn→∞Bn(ε)=0for every ε>0.\lim_{n\to\infty}B_n(\varepsilon)=0\quad\text{for every }\varepsilon>0.(5)✦ For a real ε>0\varepsilon>0, all but a uniformly vanishing fraction of pins determine at least n1−εn^{1-\varepsilon} distinct nonzero distances: Bn(ε)→0(n→∞).B_n(\varepsilon)\to 0\quad(n\to\infty).(6)✦ Let ε>0\varepsilon>0. Then there is an NN such that every configuration PP of n≥Nn\ge N points has a pin x∈Px\in P with |Dx(P)|≥n1−ε.|D_x(P)|\ge n^{1-\varepsilon}.(7)✦

In the paper, the result reads: for every fixed ε>0\varepsilon>0, sup⁡|P|=n|{x∈P:|Dx(P)|<n1−ε}|/n→0\sup_{|P|=n}|\{x\in P : |D_x(P)|<n^{1-\varepsilon}\}|/n\to0 as n→∞n\to\infty, and in particular every sufficiently large nn-point planar set has a pin determining at least n1−εn^{1-\varepsilon} distinct nonzero distances.

What Is Checked

The claim is stated over configurations in the Euclidean plane ℝ2\mathbb{R}^2, through their distance sets (1) and the extremal bad fraction (2).

The library proves, about these constructions, that for a configuration PP and a point xx of the plane, every element rr of the distance set Dx(P)D_x(P) satisfies r>0.r>0.(8)✦ It also proves, for every n≥2n\ge2 and every real ε\varepsilon, a bound of the extremal bad fraction by the extremal rich-pair fraction at half the exponent: Bn(ε)≤2Fn(ε/2)+2n1−ε/2n.B_n(\varepsilon)\le 2F_n(\varepsilon/2)+\frac{2\,n^{1-\varepsilon/2}}{n}.(9)✦