EGZ constant growth in (ℤ/kℤ)ⁿ fixed k
Open Question on s((Z/kZ)^n) for fixed k · arXiv:1708.09100
Status open high confidence
The open question of estimating s((Z/kZ)^n) for fixed k and large n remains unresolved. For prime k=p, Fox–Sauermann (2018) established s(F_p^n) ≤ 2p·r(F_p^n), giving an exponential upper bound via Ellenberg–Gijswijt, but a gap between this and Elsholtz's exponential lower bound (base ~2.08 for odd k≥3) remains open. For composite k the Alon–Dubiner bound (cn log n)^n·k remains the state of the art, far exceeding the exponential lower bound. No subsequent verified paper closes this gap.
Reviewer notes. No verified post-2018 paper found that resolves or substantially narrows the gap between the exponential lower bound (base ~2.08 for odd k≥3, Elsholtz) and the upper bounds for s((Z/kZ)^n) with fixed k and large n. Naslund (arXiv:1701.04942) proved exponential upper bounds for prime-order groups before the Fox–Sauermann paper; Hegedüs (arXiv:1712.00228, 2017) extended these to arbitrary finite abelian groups; both arXiv versions predate Fox–Sauermann and are not post-2018 results. The question for composite fixed k remains fully open, with a large gap between polynomial-exponential upper bounds and exponential lower bounds.
Context
Alon and Dubiner proved $s((\mathbb{Z}/k\mathbb{Z})^n) \leq (cn\log n)^n k$ for some absolute constant $c$, showing that for any fixed $n$ the quantity grows linearly in $k$. Elsholtz obtained lower bounds $s((\mathbb{Z}/k\mathbb{Z})^n) \geq 2.08^n$ for odd $k\geq 3$ and large $n$, but a large gap remains between these bounds and the upper bounds of Corollary 2 for fixed odd $k\geq 3$ and large $n$.
Notes. Stated as a general open problem in the field with no labeled environment; no specific attribution to prior work given.
Source paper
Erdős-Ginzburg-Ziv constants by avoiding three-term arithmetic progressions
Jacob Fox, Lisa Sauermann · 2018-04-17
https://arxiv.org/abs/1708.09100
PDF source