arXiv:2609.30225
Variational Multi-Gaussian Quantum Trajectories
Zejian Li, Jacopo Tosca
Abstract
We propose a variational framework for the efficient simulation of stochastic quantum dynamics in interacting bosonic systems. Our method generalizes the time-dependent variational principle to accommodate quantum state diffusion processes and is tailored for describing quantum trajectories unraveled from a master equation. We adopt a wavefunction ansatz composed of a coherent superposition of Gaussian wavepackets, which allows the fully analytical assembly of the variational equations of motion while capturing both the quantum trajectory dynamics and the Lindblad evolution beyond semiclassical approximations. The method is carefully benchmarked on a Bose-Hubbard dimer, and applied to extended Bose-Hubbard lattices in both 1D and 2D. In particular, we show the emergence of a symmetry-breaking phase transition in 2D lattices, which is absent in 1D chains.
AI Summary
The paper addresses simulation of continuously monitored, interacting bosonic systems beyond single-Gaussian and semiclassical approximations. It develops a stochastic time-dependent variational principle by converting heterodyne quantum-state diffusion to Stratonovich form and projecting the dynamics onto a variational tangent space. The wavefunction is a coherent superposition of $N_c$ Gaussian wavepackets, with optional low-rank covariance matrices. For polynomial Hamiltonians and loss operators, Gaussian integration gives analytical expressions for the variational metric and forces. Benchmarks use a driven-dissipative Bose–Hubbard dimer with single-particle loss. Four Gaussian components reproduce an example exact trajectory, while six reproduce trajectory-averaged antibonding populations and Rényi-2 entanglement across a drive sweep, including regimes where truncated Wigner and single-Gaussian trajectories deviate. Increasing $N_c$ reduces the trajectory error. Under the large-occupation scaling $F\to F\sqrt{\mathcal N}$ and $U\to U/\mathcal N$, results approach the previously studied symmetry-breaking and trajectory-entanglement transitions. Applications to periodic chains and square lattices find decreasing staggered-mode population per site with increasing size in 1D, but persistent checkerboard order in 2D, suggesting a symmetry-breaking phase only in 2D. The largest reported 2D lattice is $6\times6$. Lattice calculations use up to $N_c=4$ and covariance rank $K=2$; convergence is checked against $N_c$ at one drive near the ordering maximum, with $K$ fixed. The thermodynamic phase-transition conclusion is therefore presented as suggestive rather than established by critical scaling. Higher $N_c$ and $K$ are expected to be necessary for converged lattice entanglement measures.
Central claim may not hold: The abstract states a dimension-dependent thermodynamic phase transition more strongly than the finite-size lattice evidence establishes. The reported lattice convergence checks do not test covariance rank or establish convergence across the drive interval used to infer the phase. The central stochastic TDVP equation incorrectly replaces the inverse of the real part of the tangent-space Gram matrix by the real part of its inverse, so it does not generally implement the stated variational projection.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 7 pages, 4 figures