Contemporary tensor network approaches to gapless and topological phases in an extended Bose-Hubbard ladder

The development of numerically efficient computational methods has facilitated in depth studies of various correlated phases of matter including critical and topological phases. A quantum Monte-Carlo study of an extended Bose-Hubbard ladder has recently been used to identify an exotic phase with hidden order, where superfluid correlations coexist with string order, dubbed a Haldane superfluid (HSF). However, finite-size methods can struggle to uniquely determine the boundaries of quasi-long-range ordered states with nonlocal, e.g. string-like, correlations. In the present Letter, we revisit the HSF scenario using tensor network algorithms specialized for finite/infinite (quasi-)1D systems, \textit{i.e.} the well-governed finite-size density matrix renormalization group (DMRG), and the state-of-the-art infinite-size variational uniform matrix product state (VUMPS) methods. While DMRG results extrapolated to the thermodynamic limit are compatible with a putative HSF, the results from the VUMPS calculations provide sharper phase boundaries that leave no room for such a topological superfluid. Our results demonstrate the crucial advantage of the VUMPS in characterizing topological and critical interacting phases providing the precise phase boundaries.