Convex graphon parameters and graph norms

TL;DR

This paper investigates the weakly norming property of certain bipartite graphs, proving that specific complex graphs like $K_{5,5}\setminus C_{10}$ are not weakly norming, advancing understanding of graph norms and Sidorenko's conjecture.

Contribution

It demonstrates that twisted blow-ups of cycles, including $K_{5,5}\setminus C_{10}$, are not weakly norming, and shows that $K_{t,t}$ minus a perfect matching is not norming for all $t>3$, answering open questions.

Findings

$K_{5,5}\setminus C_{10}$ is not weakly norming.

Twisted blow-ups of cycles are not weakly norming.

$K_{t,t}$ minus a perfect matching is not norming for $t>3$.

Abstract

Sidorenko's conjecture states that the number of copies of a bipartite graph $H$ in a graph $G$ is asymptotically minimised when $G$ is a quasirandom graph. A notorious example where this conjecture remains open is when $H=K_{5,5}\setminus C_{10}$. It was even unknown whether this graph possesses the strictly stronger, weakly norming property. We take a step towards understanding the graph $K_{5,5}\setminus C_{10}$ by proving that it is not weakly norming. More generally, we show that 'twisted' blow-ups of cycles, which include $K_{5,5}\setminus C_{10}$ and $C_6\square K_2$, are not weakly norming. This answers two questions of Hatami. The method relies on the analysis of Hessian matrices defined by graph homomorphisms, by using the equivalence between the (weakly) norming property and convexity of graph homomorphism densities. We also prove that $K_{t,t}$ minus a perfect matching, proven to be weakly norming by Lov\'asz, is not norming for every $t>3$.