Squeezing is known to be a quantum resource in many applications in metrology, cryptography, and computing, being related to entanglement in multimode settings. In this work, we address the effects of squeezing in neuromorphic machine learning for time series processing. In particular, we consider a loop-based photonic architecture for reservoir computing and address the effect of squeezing in the reservoir, considering a Hamiltonian with both active and passive coupling terms. Interestingly, squeezing can be either detrimental or beneficial for quantum reservoir computing when moving from ideal to realistic models, accounting for experimental noise. We demonstrate that multimode squeezing enhances its accessible memory, which improves the performance in several benchmark temporal tasks. The origin of this improvement is traced back to the robustness of the reservoir to readout noise as squeezing increases.

We consider a class of open quantum many-body systems that evolves in a Markovian fashion, the dynamical generator being in GKS-Lindblad form. Here, the Hamiltonian contribution is characterized by an all-to-all coupling, and the dissipation features local transitions that depend on collective, operator-valued rates, encoding average properties of the system. These types of generators can be formally obtained by generalizing, to the quantum realm, classical (mean-field) stochastic Markov dynamics, with state-dependent transitions. Focusing on the dynamics emerging in the limit of infinitely large systems, we build on the exactness of the mean-field equations for the dynamics of average operators. In this framework, we derive the dynamics of quantum fluctuation operators, that can be used in turn to understand the fate of quantum correlations in the system. We apply our results to quantum generalized Hopfield associative memories, showing that, asymptotically and at the mesoscopic scale only a very weak amount of quantum correlations, in the form of quantum discord, emerges beyond classical correlations.