Searcharxiv⌕ Search

arXiv · 2610.34038

Extremal 3-uniform hypergraphs without generalized 4-cycles

Abstract

Let $f_r(n)$ be the maximum number of edges in an $n$-vertex $r$-uniform hypergraph in which all disjoint pairs of edges have distinct unions. In 1977, Erdős posed the problem of determining $f_r(n)$ and, in particular, asked whether $f_3(n)=O(n^2)$. Füredi later proved that $\binom{n-1}{r-1}+\left\lfloor (n-1)/r\right\rfloor \le f_r(n)<\frac72\binom{n}{r-1}$ for $r\ge3$ and $n\ge2r$. Motivated by his constructions, he further conjectured that $f_3(n)\le\binom{n}{2}$ for all sufficiently large $n$, and that, for each fixed $r\ge4$, the lower bound is exact for all sufficiently large $n$. This is known as the Erdős--Füredi conjecture. In a recent breakthrough, Huang, Ma and Yang proved the conjecture for every fixed $r\ge4$ and characterized all extremal hypergraphs. In this paper, we settle the remaining $3$-uniform case for every $n$. The bound is sharp whenever $n\ge5$ and $n\equiv1,5\pmod{20}$. Our proof uses a local weight inequality and a transfer argument.

Explore related subjects

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Leilei Zhang. 2026-10-25. Extremal 3-uniform hypergraphs without generalized 4-cycles. https://arxiv.org/abs/2610.34038

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

A Study On The Graph Formulation Of Union Closed Sets Conjecture

The Union Closed Sets Conjecture is one of the most renowned problems in combinatorics. Its appeal lies in the simplicity of its statement contrasted with the potential complexity of its resolution. The conjecture posits that, in any union closed family of sets, there exists at least one element that appears in at least half of the sets within the family. We establish the graph-theoretic version of certain set-theoretic results by connecting the set-based and graph-based formulations. We then prove a theorem in which we investigate the conjecture for graphs, focusing on their decompositions and the position of certain pendant vertices. As a result, we extend the validity of the conjecture to a broader class of graph structures.

math.CO↗

Embedding edge-colored graphs in expanders with roll-back

We introduce a method to embed edge-colored graphs into families of expander graphs, which generalizes a framework developed by Draganić, Krivelevich, and Nenadov (2022). As an application, we show that each family of sufficiently pseudo-random graphs on $n$ vertices contains every edge-colored subdivision of $K_Δ$, provided that the distance between branch vertices in the subdivision is large enough, the average degree of each graph in the family is at least $(1+o(1))Δ$, and the number of vertices in the subdivision is at most $(1-o(1))n$. This extends the subdivision-embedding results of Draganić, Krivelevich, and Nenadov to the colorful setting. Even in the monochromatic case, our work extends the earlier result, which only applied to graphs with average degree at most $O(n^{1/5})$. This work is motivated in part by the problem of finding structures in distance graphs defined over finite vector spaces. In particular, we obtain nearly spanning subdivisions of complete graphs with arbitrary prescribed nonzero edge distances in sufficiently large subsets of these spaces.

math.CO↗

Identically vanishing $k$-generalized Fibonacci polynomials

The recurrence for the $k$-Fibonacci polynomials is usually iterated upwards to positive values of $n$ only. When the recurrence is iterated downwards to $n<0$, there are indices where the polynomials vanish identically. This fact does not seem to have been noted in the literature. We derive the set of such indices. We establish the connection of our results to the solution of the Skolem problem for the $k$-Fibonacci numbers. For $k\ge3$ and $n<0$, we show that the degree of the polynomial does not increase monotonically with $|n|$. The so-called `left-justified $k$-nomial triangle' is extended to treat negative indices. We derive expressions for the individual polynomial coefficients (the elementary symmetric polynomials of the roots). We present results for the properties of the polynomials, for both $n>0$ and $n<0$, including factorization of the polynomials and properties of the roots. Results are also derived for real roots. We present new, tighter, bounds on the amplitudes of the nonzero roots. We derive new combinatorial sums for the polynomial coefficients, which are more concise and computationally efficient than previously published expressions.

math.CO↗