Efficient Application of Tensor Network Operators to Tensor Network States Through Successive Deterministic Compression
This work introduces Cholesky-Based Compression (CBC), a novel algorithm for efficiently applying tree tensor network operators to states. CBC splits tensor networks, uses Cholesky decomposition on matrix products to manage bond dimensions, and performs sweeps for both tensor train and general tree structures. It achieves state-of-the-art performance with at least an order of magnitude faster runtime.