Pair-complexity lower bound for 3-graph Ramsey
Conjecture 4.1 · arXiv:2404.02021
Status open medium confidence
Conjecture 4.1 from arXiv:2404.02021 asks whether every Ramsey coloring of order N for 3-graphs G,H with pair-construction complexity p > r(G,H)^{o(1)} must have complexity at least r(G,H)^a for an absolute constant a>0. No follow-up paper confirming or refuting this conjecture was found. A related 2026 preprint by He (arXiv:2603.16069) constructs 3-graphs with quasipolynomial Ramsey growth rates in the same research area, but its abstract does not address the pair-construction complexity framework of Conjecture 4.1.
Reviewer notes. The conjecture is a structural/framework-level claim asserting that pair constructions of low complexity are essentially universal for hypergraph Ramsey lower bounds. The source paper was published in IMRN (Volume 2025, Issue 11). A 2026 preprint arXiv:2603.16069 by He (a co-author) constructs 3-graphs with quasipolynomial Ramsey growth rates n^{Theta(log n)}, directly relevant to the surrounding discussion, but could not be confirmed to address Conjecture 4.1 specifically within the 5-call budget.
Context
The authors introduce the notion of pair constructions $\chi_{f,g}$ of complexity $p$, observing that stepping-up and inducibility constructions both fit this paradigm. They conjecture that any Ramsey coloring of order $N$ yields one of order $N^{\Omega(1)}$ with much smaller complexity, which would explain the prevalence of low-complexity constructions in Ramsey theory and potentially illuminate the possible growth rates of $r(H, K_{n}^{(3)})$.
Source paper
On off-diagonal hypergraph Ramsey numbers
David Conlon, Jacob Fox, Benjamin Gunby, Xiaoyu He, Dhruv Mubayi, Andrew Suk, Jacques Verstraete · 2024-04-02
https://arxiv.org/abs/2404.02021