Title: Quasiparticle cooling algorithms for quantum many-body state preparation

Abstract: Probing correlated states of many-body systems is one of the central tasks for quantum simulators and processors. A promising approach to state preparation is to realize desired correlated states as steady states of engineered dissipative evolution. A recent experiment with a Google superconducting quantum processor [X. Mi et al., Science 383, 1332 (2024)] demonstrated a cooling algorithm utilizing auxiliary degrees of freedom that are periodically reset to remove quasiparticles from the system, thereby driving it towards the ground state. We develop a kinetic theory framework to describe quasiparticle cooling dynamics, and employ it to compare the efficiency of different cooling algorithms. In particular, we introduce a protocol where coupling to auxiliaries is modulated in time to minimize heating processes, and demonstrate that it allows a high-fidelity preparation of ground states in different quantum phases. We verify the validity of the kinetic theory description by an extensive comparison with numerical simulations of a 1d transverse-field Ising model using a solvable model and tensor-network techniques. Further, the effect of noise, which limits efficiency of variational quantum algorithms in near-term quantum processors, can be naturally described within the kinetic theory. We investigate the steady state quasiparticle population as a function of noise strength, and establish maximum noise values for achieving high-fidelity ground states. This work establishes quasiparticle cooling algorithms as a practical, robust method for many-body state preparation on near-term quantum processors.