Accelerating quantum embedding beyond DFT using interpolative separable density fitting

Beyond the need for efficient solvers for the strongly correlated subspace, ab initio quantum embedding beyond density functional theory (DFT) requires a scalable many-body treatment of the weakly correlated environment. In this talk, I will present our recent development of low-scaling algorithms for many-body electronic structure based on interpolative separable density fitting (ISDF). ISDF is a systematically controlled compression technique for general many-body Hamiltonians in any canonical basis, providing a representation that achieves separability in orbital and k-point indices. This separability enables the construction of low-scaling algorithms for subsequent many-body calculations, with applications including GW, vertex corrections, and extensions of dynamical mean-field theory.