Optimal storage capacity of quantum Hopfield neural networks

Quantum neural networks form one pillar of the emergent field quantum machine learning. Here generalizations classical realizing associative memories---capable retrieving patterns, or memories, from corrupted initial states---have been proposed. It is a challenging open problem to analyze memories with an extensive number patterns and determine maximal can reliably store, i.e., their storage capacity. In this work, we propose explore general method for evaluating capacity network models. By generalizing what known as Gardner's approach in realm, exploit theory spin glasses deriving optimal quenched pattern variables. As example, apply our open-system memory formed interacting spin-1/2 particles coupled artificial neurons. The system undergoes Markovian time evolution resulting dissipative retrieval dynamics that competes coherent dynamics. We map out nonequilibrium phase diagram study effect temperature Hamiltonian on Our opens avenue systematic characterization memories.