arXiv++ Combinatorics

We develop a topological framework for proving lower bounds on sign-rank via $\mathbb{Z}_2$-equivariant topology, and use it to resolve the sign-rank of the Gap Hamming Distance problem up to lower-order terms. For every (partial) sign matrix $A$, we associate a free $\mathbb{Z}_2$-simplicial complex $S(A)$ and show that sign-rank of $A$ is characterized by the linear analog of $\mathbb{Z}_2$-index of $S(A)$. As a consequence, the classical $\mathbb{Z}_2$-index of $S(A)$ lower bounds the sign-rank of $A$, which reduces sign-rank lower bounds to topological obstructions. This reduction allows us to use various tools from $\mathbb{Z}_2$-equivariant topology, particularly in regimes where classical lower-bound techniques break down. As the main application, we consider the Gap Hamming Distance function $\mathrm{GHD}_k^n$ (defined for $k < n/2$), which distinguishes pairs of strings in $\{0,1\}^n$ with Hamming distance at most $k$ from pairs with distance at least $n-k$. We prove an essentially tight lower bound and show that for any $k$, \[ \text{sign-rank}(\mathrm{GHD}_k^n) = (1-o_k(1)) 2k. \] where the $o_k(1)$ term is $O\left(\sqrt{\frac{\log k}{k}}\right)$. This improves on the previous lower bound of Hatami, Hosseini, and Meng (STOC 2023) who proved that sign-rank of $\mathrm{GHD}_k^n$ is at least $Ω(k/\log(n/k))$. A key technical ingredient is a new analysis of the $\mathbb{Z}_2$-coindex (which lower bounds $\mathbb{Z}_2$-index) of the Vietoris-Rips complex of the hypercube in the sparse regime which yields an essentially tight lower bound. Previously, no results were known in the sparse regime.

In a recent breakthrough, Kelley and Meka (FOCS 2023) obtained a strong upper bound on the density of sets of integers without nontrivial three-term arithmetic progressions. In this work, we extend their result, establishing similar bounds for all linear patterns defined by binary systems of linear forms, where "binary" indicates that every linear form depends on exactly two variables. Prior to our work, no strong bounds were known for such systems even in the finite field model setting. A key ingredient in our proof is a graph counting lemma. The classical graph counting lemma, developed by Thomason (Random Graphs 1985) and Chung, Graham, and Wilson (Combinatorica 1989), is a fundamental tool in combinatorics. For a fixed graph $H$, it states that the number of copies of $H$ in a pseudorandom graph $G$ is similar to the number of copies of $H$ in a purely random graph with the same edge density as $G$. However, this lemma is only non-trivial when $G$ is a dense graph. In this work, we prove a graph counting lemma that is also effective when $G$ is sparse. Moreover, our lemma is well-suited for density increment arguments in additive number theory. As an immediate application, we obtain a strong bound for the Turán problem in abelian Cayley sum graphs: let $Γ$ be a finite abelian group with odd order. If a Cayley sum graph on $Γ$ does not contain any $r$-clique as a subgraph, it must have at most $2^{-Ω_r(\log^{1/16}|Γ|)}\cdot |Γ|^2$ edges. These results hinge on the technology developed by Kelley and Meka and the follow-up work by Kelley, Lovett, and Meka (STOC 2024).

The Bogolyubov-Ruzsa lemma, in particular the quantitative bounds obtained by Sanders, plays a central role in obtaining effective bounds for the inverse $U^3$ theorem for the Gowers norms. Recently, Gowers and Milićević applied a bilinear Bogolyubov-Ruzsa lemma as part of a proof of the inverse $U^4$ theorem with effective bounds. The goal of this note is to obtain quantitative bounds for the bilinear Bogolyubov-Ruzsa lemma which are similar to those obtained by Sanders for the Bogolyubov-Ruzsa lemma. We show that if a set $A \subset \mathbb{F}_p^n \times \mathbb{F}_p^n$ has density $α$, then after a constant number of horizontal and vertical sums, the set $A$ would contain a bilinear structure of co-dimension $r=\log^{O(1)} α^{-1}$. This improves the results of Gowers and Milićević which obtained similar results with a weaker bound of $r=\exp(\exp(\log^{O(1)} α^{-1}))$ and by Bienvenu and Lê which obtained $r=\exp(\exp(\exp(\log^{O(1)} α^{-1})))$.

Let $G$ be a finite abelian group and $A$ a subset of $G$. The spectrum of $A$ is the set of its large Fourier coefficients. Known combinatorial results on the structure of spectrum, such as Chang's theorem, become trivial in the regime $|A| = |G|^α$ whenever $α\le c$, where $c \ge 1/2$ is some absolute constant. On the other hand, there are statistical results, which apply only to a noticeable fraction of the elements, which give nontrivial bounds even to much smaller sets. One such theorem (due to Bourgain) goes as follows. For a noticeable fraction of pairs $γ_1,γ_2 $ in the spectrum, $γ_1+γ_2$ belongs to the spectrum of the same set with a smaller threshold. Here we show that this result can be made combinatorial by restricting to a large subset. That is, we show that for any set $A$ there exists a large subset $A'$, such that the sumset of the spectrum of $A'$ has bounded size. Our results apply to sets of size $|A| = |G|^α$ for any constant $α>0$, and even in some sub-constant regime.

The arithmetic regularity lemma due to Green [GAFA 2005] is an analogue of the famous Szemer{é}di regularity lemma in graph theory. It shows that for any abelian group $G$ and any bounded function $f:G \to [0,1]$, there exists a subgroup $H \le G$ of bounded index such that, when restricted to most cosets of $H$, the function $f$ is pseudorandom in the sense that all its nontrivial Fourier coefficients are small. Quantitatively, if one wishes to obtain that for $1-ε$ fraction of the cosets, the nontrivial Fourier coefficients are bounded by $ε$, then Green shows that $|G/H|$ is bounded by a tower of twos of height $1/ε^3$. He also gives an example showing that a tower of height $Ω(\log 1/ε)$ is necessary. Here, we give an improved example, showing that a tower of height $Ω(1/ε)$ is necessary.