On sums and products of combinatorial cubes (Q2667086)

Let \(a_0,a_1, \ldots,a_d\) be fixed non-zero real numbers. A combinatorial cube (also known as a Hilbert cube) is a set of the form \[ Q(a_0,a_1,\ldots,a_d ):= \left \{ a_0 + \varepsilon_1 a_1 + \varepsilon_2a_2 + \cdots + \varepsilon_d a_d : \varepsilon_1,\dots, \varepsilon_d \in \{0,1 \} \right\}. \] Since this set is defined additively, the principles behind sum-product theory lead us to believe that combinatorial cubes should not exhibit multiplicative structure. To be more precise, one may expect that, for a combinatorial cube \(Q\), the product set \(QQ\) is guaranteed to be large. A stronger result in this direction would be to show that the multiplicative energy of \(Q\) cannot be large, where the multiplicative energy of a set \(A\) is defined to be \[ E^*(A)=| \{ (a,b,c,d) \in A^4 : ab=cd \}|. \] This paper proves such results. In particular, it is proven that there exists a constant \(c>0\) such that, for any combinatorial cube \(Q \subset \mathbb R\), \[ E^*(Q) \leq |Q|^{3-c}. \] This immediately implies a non-trivial lower bound for the size of the product set \(QQ\) and ratio set \(Q/Q\). The paper also gives a quantitatively stronger bound for both sets, proving that \[ |QQ| \gg |Q|^{100/79}, \quad |Q/Q| \gg |Q|^{14/11}. \] Slightly weaker bounds for the product and ratio sets of combinatorial cubes over \(\mathbb F_p\) are also given. The paper also proves dual versions of all of the results mentioned above, where the roles of addition and multiplication are reversed. In particular, one can define a multiplicative analogue of a combinatorial cube as follows: \[ Q^*(a_0,a_1,\dots,a_d ):= \left \{ a_0 \cdot a_1^{\varepsilon} \cdot \cdots \cdot a_d^{\varepsilon_d} : \varepsilon_1,\dots, \varepsilon_d \in \{0,1 \} \right\}. \] This paper proves a non-trivial upper bound for the additive energy of a multiplicative combinatorial cube, and gives quantitatively better lower bounds for the size of its sum set and difference set. The proofs combine methods from incidence geometry, higher energy techniques and the eigenvalue method, results on growth in groups, along with some additional combinatorial ideas.