WSK validity for 5-colorings of triangular lattice

Open case: WSK algorithm on the triangular lattice for q=5 · arXiv:1510.06964

arXiv Problem medium confidence— first stated 2016-09-22

Authors: Marthe Bonamy, Nicolas Bousquet, Carl Feghali, Matthew Johnson

Status partial medium confidence

The question of whether all 5-colourings of the triangular lattice (with periodic boundary conditions) form a single Kempe class — equivalently whether the WSK algorithm is ergodic for q=5 — was explicitly left open in arXiv:1510.06964. A 2022 preprint by Salas and Sokal (arXiv:2206.13126) proves WSK ergodicity for q≥5 on Eulerian triangulations of the torus satisfying certain structural conditions; since the triangular lattice is a 6-regular Eulerian triangulation of the torus, this is a significant adjacent result, but the abstract does not explicitly confirm that the standard triangular lattice satisfies the paper's conditions for q=5. No paper was found that unambiguously and completely settles the conjecture.

Cited literature (1)

partial Ergodicity of the Wang--Swendsen--Kotecký algorithm on several classes of lattices on the torus (2022)

Jesús Salas, Alan D. Sokal · arXiv preprint · arXiv:2206.13126

Proves WSK ergodicity for q≥5 on Eulerian triangulations of the torus meeting certain structural conditions, an adjacent result that may encompass the standard triangular lattice but does not explicitly confirm it.

Reviewer notes. arXiv:2206.13126 (Salas--Sokal 2022) is the most relevant post-2016 paper found; it establishes WSK ergodicity for q≥5 on Eulerian triangulations under certain conditions. The triangular lattice is 6-regular and Eulerian, making it a natural candidate, but the WebFetch of the abstract indicated the triangular lattice is not explicitly mentioned as a covered case. No separate paper resolving the q=5 triangular lattice case was found.

Auto-reviewed 2026-05-15 with claude-sonnet-4-6 (web search enabled).

Problem. Determine whether the Wang-Swendsen-Koteck\'{y} algorithm for $q = 5$ colourings of the triangular lattice (with periodic boundary conditions) is valid (i.e., whether the set of $5$-colourings of the triangular lattice is a Kempe class).

Context

Theorem 2.2 establishes that the WSK algorithm on the triangular lattice is valid for $q \geq 6$ (via Theorem 1.1 and a degeneracy lemma) and invalid for $q \leq 4$ (Mohar and Salas). The authors explicitly note that $q = 5$ is the single remaining open case.

Notes. Stated as an observation ('We observe that this leaves the single open case of a triangular lattice with q = 5') rather than in a formal labelled environment.

Source paper

On a conjecture of Mohar concerning Kempe equivalence of regular graphs
Marthe Bonamy, Nicolas Bousquet, Carl Feghali, Matthew Johnson · 2016-09-22
https://arxiv.org/abs/1510.06964 PDF source