The deletion-normality conjecture for non-complete graphs

From papers

For a graph HH, an HH-deletion-saturated graph is an HH-free graph GG with E(G)E(G)\neq\varnothing such that, for every eE(G)e\in E(G), the graph GeG-e is not HH-free. A graph HH is deletion-normal if there exists an HH-deletion-saturated graph. Likewise, HH is addition-normal if there exists an HH-free graph GG with E(G)E(\overline{G})\neq\varnothing such that G+eG+e is not HH-free for every eE(G)e\in E(\overline{G}). Deletion-normality conjecture. Every non-complete graph HH is deletion-normal. Equivalently, every non-edgeless graph HH is addition-normal. Complete graphs on at least two vertices are the only graphs currently known not to be deletion-normal, while edgeless graphs on at least two vertices are the only graphs currently known not to be addition-normal; the conjecture is proved for several classes, including complete bipartite graphs with unequal part sizes, triangle-free unicyclic graphs, graphs with two leaves at distance at most three, and line graphs of trees.

Progress summary

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Sources & referencesView supporting material

Primary source

Xinyue Fan, Sahab Hajebi, Sepehr Hajebi and Sophie Spirkl, “Asymmetric induced saturation”, arXiv:2606.24763 (2026).

Solutions 0

No solutions have been posted yet.