论文标题
光锥张量网络和时间演变
Light cone tensor network and time evolution
论文作者
论文摘要
横向折叠算法[Phys。莱特牧师。 102，240603]是一种张量网络方法，用于计算时间依赖的局部可观察物，在量子不平衡的量子旋转链中，有时可以克服矩阵乘积状态的局限性。我们提出了一种收缩策略，该策略利用代表观察力的张量网络的确切光锥结构。该策略可以与[物理学中的全局淬火所提出的混合截断结合使用。 Rev. A 91，032306]，可显着提高该方法的效率。我们证明了该横向光锥的性能也用于传输系数，并讨论如何将其扩展到其他动力学量。
The transverse folding algorithm [Phys. Rev. Lett. 102, 240603] is a tensor network method to compute time-dependent local observables in out-of-equilibrium quantum spin chains that can sometimes overcome the limitations of matrix product states. We present a contraction strategy that makes use of the exact light cone structure of the tensor network representing the observables. The strategy can be combined with the hybrid truncation proposed for global quenches in [Phys. Rev. A 91, 032306], which significantly improves the efficiency of the method. We demonstrate the performance of this transverse light cone contraction also for transport coefficients, and discuss how it can be extended to other dynamical quantities.
