Description
We present a framework that leverages Clifford-enhanced matrix product state (MPS) techniques to tackle the challenges of simulating quantum many-body dynamics. By integrating stabilizer-based disentangling strategies into MPS algorithms, we enable efficient simulations of both circuit-based and Hamiltonian-driven evolutions. Our approach begins with a hybrid method combining tensor networks and random Clifford circuits, demonstrating reduced entanglement growth and improved simulation capabilities in circuit dynamics. Building on this foundation, we develop a Clifford-dressed time-dependent variational principle (TDVP) algorithm that actively suppresses entanglement during Hamiltonian evolution, significantly extending accessible simulation times and improving accuracy in local observables. Finally, we extend this framework to compute quantities beyond standard expectation values—such as Loschmidt echoes—by exploiting overlaps between MPS and stabilizer states. This suite of techniques opens new pathways for scalable simulations of quantum systems, bridging computational efficiency with the richness of many-body physics.
References:
Hybrid Stabilizer Matrix Product Operator. AFM, A. Santini, M. Collura. PRL. (2024)
Clifford dressed time-dependent variational principle. AFM, A. Santini, G. Lami, J. De Nardis, M. Collura. PRL. (2025)
Clifford-dressed variational principles for precise Loschmidt echoes. AFM, A. Santini, M. Collura. Accepted in PRA. (2025)
