“Quantum networks are at the core of a flurry of quantum information tasks, such as quantum communication, quantum key distribution, and quantum computation, as well as the pillar of the future quantum internet. In this context, it is pivotal that users of future quantum networks are equipped with protocols for the certification of their quantum devices, possibly relying on minimal hardware assumptions (device-independently). However, standard device-independent certification methods mainly rely on semi-definite programming approaches. Such optimization methods have two main drawbacks: first, they can only deal with linear constraints and linear objective functions, and, second, they become easily unfeasible when increasing the size of quantum networks. In this talk, I will report on a machine learning-based approach for the device-independent certification of arbitrarily large and complex quantum networks. Such an approach proves more efficient than standard tools, and it can be used not only to optimize arbitrary functions over supersets of quantum correlations but also to lower-bound violations of Bell-like inequalities compatibly with experimentally observed sets of correlations. We believe that this work may represent a significant step forward toward the certification of arbitrarily large and complex quantum networks”.
Beatrice Polacchi will be talking in Paris on 31 October 2023.