Higher symplectic capacities and the stabilized embedding problem for integral elllipsoids (Q2155485)

The stabilized embedding problem is regarding the existence of a symplectic embedding \(X_1 \times \mathbb{C}^N \hookrightarrow X_2 \times \mathbb{C}^N\) for some \(N \geq 0\) where \(X_1\) and \(X_2\) are symplectic four-manifolds. Given two reals \(x\leq y\), ellipsoids and polydisks are defined as the following symplectic four-manifolds \[ E(x,y) := \left\{\frac{\pi |z_1|^2}{x}+\frac{\pi |z_2|^2}{y} \leq 1\right\} \subset \mathbb{C}^2, \qquad P(x,y) := \left\{\frac{\pi |z_1|^2}{x} \leq 1, \; \frac{\pi |z_2|^2}{y} \leq 1\right\} \subset \mathbb{C}^2. \] In the paper under review, the authors study the stabilized embedding ellipsoid-into-ellipsoid problem and the stabilized embedding ellipsoid-into-polydisk problem and solve it completely in some cases where \(x\) and \(y\) above are integers. Define \(c^N_{b,\text{ell}}(a)\) to be the infimum over \(\lambda\), such that there exists an embedding \(E(1,a) \times \mathbb{C}^N \hookrightarrow E(\lambda,\lambda b) \times \mathbb{C}^N\). Similarly we define \(c^N_{b,\text{poly}}(a)\) to be the infimum over \(\lambda\), such that there exists an embedding \(E(1,a) \times \mathbb{C}^N \hookrightarrow P(\lambda,\lambda b) \times \mathbb{C}^N\). In the paper under review, the authors prove the following. \begin{itemize} \item[1.] Let \(a\) and \(b\) be integers with different parity such that \(b\geq 1\) and \(a\geq b+1\). Then for \(N\geq 1\) \[ c^N_{b,\text{ell}}(a) = \frac{2a}{a+b-1}. \] \item[2.] Let \(b \geq 1\) be any real number and let \(a\geq 2b-1\) be an odd integer. Then for \(N \geq 1\) \[ c^N_{b,\text{poly}}(a) = \frac{2a}{a+2b-1}. \] \end{itemize} Embedding problems, stabilized or not, are often studied using symplectic capacities, see [\textit{K. Cieliebak} et al., Math. Sci. Res. Inst. Publ. 54, 1--44 (2007; Zbl 1143.53341)] and references therein. The proof of the above theorems utilizes a new sequence of ``higher'' symplectic capacities \(\mathfrak g_k\) defined by the third author, which are invariant under multiplication by \(\mathbb{C}\) [\textit{K. Siegel}, ``Higher symplectic capacities'', Preprint, \url{arXiv:1902.01490}]. The authors utilize explicit computations of \(\mathfrak g_k(E(1,a))\) and \(\mathfrak g_k(P(1,a))\) found in \emph{loc. cit.} and [\textit{D. McDuff} and \textit{K. Siegel}, ``Symplectic capacities, unperturbed curves, and convex toric domains'', Preprint, \url{arXiv:2111.00515}].