So, not too long after I figured out the answer to my question. At least in the two dimensional case, we define an arbitrary dyadic interval of the form $[a,b]$, for $a=\frac{p}{2^k}, b=\frac{p+1}{2^k}, p \in \mathbb{Z}, k\in \mathbb{Z}^+\cup \\{0\\}$. And we also define another dyadic interval of the same length: $[c,d]$ for $c=\frac{s}{2^k}, b=\frac{s+1}{2^k}, s \in \mathbb{Z}, k\in \mathbb{Z}^+\cup \\{0\\}$. It is easy to show that the $k$ values are the same.
The cartesian product of these two intervals: $[a,b]\times[c,d]$ is a dyadic square in $\mathbb{R}^2$. The higher dimensional analogs are created in precisely the same way. So a dyadic $m$-cube is a set of points specified by an $m$ dimensional cartesian product in $\mathbb{R}^m$.