Tomographic completeness and robustness of quantum reservoir networks

Quantum reservoir processing offers an option to perform quantum tomography of input objects by postprocessing quantities, obtained from local measurements, from a quantum reservoir network that has interacted with the former. We develop a method to assess a tomographic completeness criterion for arbitrary quantum reservoir architectures. Furthermore, we propose a figure of merit that quantifies their robustness against imperfections. Measured quantities from the reservoir nodes correspond to effective observables acting on the input objects, and we provide a way to retrieve them. Finally, we present examples of quantum tomography for demonstration. Our general method offers guidance in optimizing implementations of quantum reservoir processing.