Quantifying Subspace Entanglement with Geometric Measures

Xuanran Zhu, Chao Zhang, Bei Zeng

Published 2023-11-17Version 1

Determining whether a quantum subspace is entangled and quantifying its entanglement level remains a fundamental challenge in quantum information science. This paper introduces a geometric measure of $r$-bounded rank, $E_r(S)$, for a given subspace $S$. This measure, derived from the established geometric measure of entanglement, is tailored to assess the entanglement within $S$. It not only provides a benchmark for quantifying the entanglement level but also sheds light on the subspace's ability to preserve such entanglement. Utilizing non-convex optimization techniques from the domain of machine learning, our method effectively calculates $E_r(S)$ in various scenarios. Showcasing strong performance in comparison to existing hierarchical and PPT relaxation techniques, our approach is notable for its accuracy, computational efficiency, and wide-ranging applicability. This versatile and effective tool paves the way for numerous new applications in quantum information science. It is particularly useful in validating highly entangled subspaces in bipartite systems, determining the border rank of multipartite states, and identifying genuinely or completely entangled subspaces. Our approach offers a fresh perspective for quantifying entanglement, while also shedding light on the intricate structure of quantum entanglement.