Ben Green's Open Problem 18 #
Reference:
- [Gr26] Ben Green's Open Problem 18
- [Au16] Austin, Tim. "Ajtai–Szemerédi theorems over quasirandom groups." Recent trends in combinatorics. Cham: Springer International Publishing, 2016. 453-484.
- [So13] Solymosi, Jozsef. "Roth-type theorems in finite groups." European Journal of Combinatorics 34.8 (2013): 1454-1458.
- [Go01] Gowers, William T. "A new proof of Szemerédi's theorem." Geometric & Functional Analysis GAFA 11.3 (2001): 465-588.
The number of triples $(x, y, g)$ in $G^3$ such that $g \neq e$, and $(x, y), (gx, y), (x, gy)$ are all in $A$. These are called "naive corners" by [Au16].
Note: the shortened formulation from [Gr26] does not mention $g \neq e$, but this is the original statement from [Au16], which ensure non-trivial corners. Note however that [Au16] use more generally compact groups and not just finite discrete groups.
Equations
Instances For
Suppose that $G$ is a finite group, and let $A \subset G \times G$ be a subset of density $\alpha$. Is it true that there are $\gg_\alpha |G|^3$ triples $x, y, g$ such that $(x, y), (gx, y), (x, gy)$ all lie in $A$?
Note: A is taken as $\alpha$-dense, i.e. $|A| \ge \alpha |G|^2$ [Au16, Question 2]
The number of triples $(x, y, g)$ in $G^3$ such that $g \neq e$, and $(x, y), (xg, y), (x, gy)$ are all in $A$. These are called "BMZ corners" by [Au16].
Equations
Instances For
[So13] proved this is true for "BMZ corners". Follows from the proof of Theorem 2.1, p.1456-1457.