Polynomial χ-bounding in Gyárfás-Sumner
Polynomial bounds for the Gyárfás-Sumner conjecture · arXiv:2302.08922
Status partial high confidence
The conjecture that the Gyárfás-Sumner conjecture holds with polynomial χ-bounding functions (chromatic number bounded by a polynomial in the clique number t) remains open in full generality. Significant partial progress has been made since 2023: polynomial χ-boundedness was established for P₅-free graphs (arXiv:2512.24907, Dec 2024), and the polynomial version of the conjecture was proved for intersection graphs of axis-aligned boxes of any fixed dimension (arXiv:2407.16882, Jul 2024). A further 2025 preprint (arXiv:2504.21127) improves bounds for P₅-free graphs and gives compositional results, but the full polynomial conjecture for all trees is still open.
Cited literature (3)
-
Proves the polynomial Gyárfás-Sumner conjecture for intersection graphs of axis-aligned boxes in ℝ^d (graphs of bounded boxicity): for every positive integer d and forest F, such graphs with no induced F are polynomially χ-bounded.
-
Proves that the chromatic number is polynomially bounded by the clique number for graphs with no induced five-vertex path P₅, resolving this case of the polynomial Gyárfás-Sumner conjecture via a chromatic density framework.
-
Improves the best known bound for P₅-free graphs from χ(G) ≤ ω^(log ω) to χ(G) ≤ ω^(O(log ω / log log ω)), and establishes that if T and a broom each satisfy the polynomial Gyárfás-Sumner conjecture then so does their disjoint union.
Reviewer notes. Esperet's broader conjecture that every χ-bounded hereditary class is polynomially χ-bounded was disproved by Briański, Davies and Walczak, but that does not affect this specific conjecture about forest-free graphs. The most notable advance is arXiv:2512.24907 (Dec 2024) proving polynomial χ-boundedness for P₅-free graphs. Author names for arXiv:2407.16882 and arXiv:2512.24907 were not returned by WebFetch and are marked accordingly.
Context
The authors note that polynomial bounds on the $\chi$-bounding function have recently been established for trees not containing $P_5$ as an induced subgraph, and suggest that the full Gyárfás-Sumner conjecture may admit a polynomial $\chi$-bounding function in $t$ for every fixed tree $T$.
Notes. Stated in prose as 'it is possible that…' without a labelled conjecture environment.
Source paper
A note on the Gyárfás-Sumner conjecture
Tung Nguyen, Alex Scott, Paul Seymour · 2023-02-17
https://arxiv.org/abs/2302.08922
PDF source