“…Research related to reduced Lagrangian representations that enable time-varying vector fields has advanced along multiple axes. These include in situ sampling techniques ( [2,45,51,53]), post hoc reconstruction strategies ( [3, 8, 10, 26]), theoretical and empirical error analysis ( [9, 29, 49]), feature extraction ( [16,17,20,31,54]), and application to various domains ( [42,52,57]). In this paper, we study the use of deep learning to perform post hoc reconstruction.…”