Quantum fields and machine learning

Abstract

In this thesis, we use detector models to study various properties of quantum fields. One such property is the correlations present in fields. It is known that two uncorrelated detectors, upon coupling to a quantum field, can become correlated, i.e. they harvest correlations from the field. In this work, we study the effect of the presence of extra detectors in correlation harvesting protocols. Our first main result is that a single interloper detector can sabotage the harvesting of classical and quantum correlations.

The second main result in this thesis is that machines can learn to extract different features of a quantum field by processing the outcomes of local probes. As proof-of-principle, we show how a neural network can distinguish a field's boundary conditions, predict the temperature of a field and of how it can distinguish between a Fock state and a phase-averaged coherent state.