Efficiently measuring and tuning quantum devices using machine learning

Fulfilling the promise of quantum technologies requires to be able to measure and tune several devices; fault-tolerant factorization using a surface code will require ~10⁸ physical qubits. A long-term approach, based on the success of integrated circuits, is to use electron spins in semiconducting devices. A major obstacle to creating large circuits in this platform is device variability. It is very time consuming to fully characterize and tune each of these devices and this task will rapidly become intractable for humans without the aid of automation.
I will present efficient measurements on a single quantum dot performed by a machine learning algorithm. This algorithm employs a probabilistic deep-generative model, capable of generating multiple full-resolution reconstructions from scattered partial measurements. Information theory is then used to select the most informative measurements to perform next. The algorithm outperforms standard grid scan techniques in different measurement configurations, reducing the number of measurements required by up to 4 times.
I will also show the use of Bayesian optimisation to tune a single quantum dot device. By generating a score function, we can make the algorithm find the operating regime of a device. We tune the device to the single-electron tunnelling regime searching in a high-dimensionality parameter space in less than a thousandth part of the time that it requires manually.