Maximum symmetric intersecting family in [k]^n
Size of symmetric intersecting families in $[k]^n$ · arXiv:1909.11578
Status open high confidence
Eberhard, Kahn, Narayanan, and Spirkl (arXiv:1909.11578) proved that any symmetric intersecting $A \subset [k]^n$ has size $o(k^n)$ for fixed $k \geq 3$, but the precise asymptotics remain open. The best known constructions, based on set-intersecting families derived from projective planes $PG(2,q)$, achieve size roughly $k^{n-\sqrt{n}}$. No matching upper bound or improvement to the construction is known, and no follow-up resolving the question was found in the indexed literature.
Reviewer notes. No follow-up resolving the precise asymptotics of the maximum size of a symmetric intersecting family in $[k]^n$ was found. The gap between the known upper bound $o(k^n)$ and the best construction $k^{n-\sqrt{n}}$ remains wide. The Keevash 2023 paper on 'forbidden intersections for codes' appeared in search results but was behind a paywall; it may or may not address this specific question.
Context
Theorem 1.1 establishes that any symmetric intersecting $A \subset [k]^n$ has size $o(k^n)$, but the precise asymptotics remain open. The best constructions known to the authors are set-intersecting families, e.g., those based on lines of a projective plane $PG(2,q)$, yielding size roughly $k^{n-\sqrt{n}}$.
Notes. PDF source — math notation may be garbled. No labeled environment; described by the authors as 'the most obvious question raised by the present work' in Section 4.
Source paper
On symmetric intersecting families of vectors
Sean Eberhard, Jeff Kahn, Bhargav Narayanan, Sophie Spirkl · 2020-07-31
https://arxiv.org/abs/1909.11578
PDF source