Forest-filleting of coherent ideals

Conjecture 2.4 · arXiv:1804.01060

arXiv Conjecture high confidence— first stated 2020-06-02

Authors: Maria Chudnovsky, Alex Scott, Paul Seymour, Sophie Spirkl

Status open high confidence

Conjecture 2.4 from arXiv:1804.01060 states that for every graph H and every forest P of H, every coherent ideal contains a P-filleting of H. The paper's main Theorem 2.3 establishes the case where P is a path of H; the conjecture extends this to all forests. No follow-up paper confirming or disproving the full conjecture was found in a targeted web search covering literature through 2026. The sole internal reference (arXiv:2508.14332) concerns a counterexample to the weak coarse Menger conjecture and is unrelated to the P-filleting conjecture.

Reviewer notes. No follow-up resolving or disproving the conjecture was found after five web calls. The conjecture is a strengthening of Theorem 2.3 (proven for P a path); the authors note it is best possible since the conclusion fails when P contains a cycle of H. It remains open as of the review date.

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

Conjecture. Let $H$ be a graph and let $P$ be a forest of $H$. Then every coherent ideal contains a $P$-filleting of $H$.

Context

The paper's main theorem (2.3) proves the statement when $P$ is a path of $H$. The authors propose this strengthening, noting it would be best possible since the conclusion fails when $P$ contains a cycle of $H$ (as coherent ideals can have arbitrarily large girth). If true, it would imply the main theorem of [8] (strong Erdős-Hajnal property for all forests) in the same way Theorem 2.3 implies the main theorem of [18].

Source paper

Pure pairs. II. Excluding all subdivisions of a graph
Maria Chudnovsky, Alex Scott, Paul Seymour, Sophie Spirkl · 2020-06-02
https://arxiv.org/abs/1804.01060 PDF source