Absorbing-State Transitions in Interactive Quantum Dynamics

Interactive quantum many-body dynamics combines unitary evolution, controlled measurements, and feedback conditioned on measurement outcomes. Such dynamics arises naturally in various contexts, from quantum error correction to state-preparation protocols, and are the natural mode of operation of quantum simulators. By tuning measurement and feedback rates, one can drive a dynamical phase transition between an "absorbing" phase in which the feedback can efficiently steer the system towards a desired many-body target state, and an "active" phase in feedback is not effective. I will discuss properties of such absorbing-state transitions in a few different contexts, focusing both on the correlations in the target state and the degree of non-locality in the feedback operations. A natural setup is dynamics in which excitations above a target state annihilate pairwise with increasing spatial range. By tuning the degree of non-locality, we uncover a line of new absorbing-state universality classes with exponents that continuously interpolate between the canonical examples of directed-percolation and parity-conserving universality classes. Our results are obtained by mapping the trajectory-averaged dynamics to a classical stochastic process with diffusion, branching and long-rang annihilation, and are of independent interest in the study of such ubiquitous classical processes.

References:

[1] Phys. Rev. B 109, L020304 (2024),

[2] arXiv:2409.03280