Rainbow stacking universality for odd Kₙ
Question 3.3 · arXiv:2405.14795
Status open high confidence
Question 3.3 asks whether every pair of proper edge-colorings of $K_n$ admits a rainbow stacking when $n$ is odd; this is known affirmatively for $n=3$ (by inspection) and $n=5$ (by computer search), contrasting with the even case where Proposition 3.2 of the source paper gives obstructing constructions via Latin squares over $\mathbb{F}_2^k$. The source paper was published in the Bulletin of the London Mathematical Society in 2025, but this is the journal publication of the same arXiv preprint, not a follow-up resolving the question. No subsequent paper resolving the general odd-$n$ case was found in a wide web search.
Reviewer notes. No follow-up resolving Question 3.3 found. The paper appeared in Bulletin of the London Mathematical Society 57(6):1656-1670 (2025), which is the journal version of the arXiv preprint and does not resolve the question. The question is expected to be difficult according to the authors. Web search and fetch of the HTML version confirmed the conjecture remains open as of the search date.
Context
Proposition 3.2 constructs, for each even $n\geq 4$, a pair of proper edge-colorings of $K_n$ with no rainbow stacking (based on a construction from transversals in Latin squares over $\mathbb{F}_2^k$). The question asks whether the odd case is categorically different. The answer is known to be affirmative for $n=3$ (by inspection) and $n=5$ (by computer search); a general resolution is expected to be difficult.
Source paper
Rainbow Stackings of Random Edge-Colorings
Noga Alon, Colin Defant, Noah Kravitz · 2024-05-23
https://arxiv.org/abs/2405.14795