Autonomous Tuning and Charge-State Detection of Gate-Defined Quantum Dots

Theoretical Bounds on Data Requirements for the Ray-Based Classification

“…Using variational autoencoders, standard device measurements have been optimized to reduce the total number of measurements required [24] and to automate fine tuning in higher ( N > 2 ) dimensions [25]. Machine learning-based binary classifiers have been used to classify 2D stability diagrams as either good or bad for further experimental use [26]. Several different CNNs have been implemented to classify 2D dot data using experimental [27], simulated [28], or a combination of both data types [29].…”

Machine learning enables completely automatic tuning of a quantum device faster than human experts

“…Moreover, the transport features that indicate the device is tuned as a double quantum dot are time-consuming to measure and difficult to parametrise. Machine learning techniques and other automated approaches have been used for tuning quantum devices [5][6][7][8][9][10][11][12][13][14] . These techniques are limited to small regions of the device parameter space or require information about the device characteristics.…”

Colloquium: Advances in automation of quantum dot devices control

“…One way to achieve this goal in the context of gate-defined QD devices is to combine physics-informed fitting and thresholds to inform the selection or relative voltage ranges as well as consecutive adjustments (Baart et al, 2016;McJunkin, 2021). An approach to characterizing gates involving fitting and binary classification has also been proposed (Darulová et al, 2020). Here, a set of parameters defining a hyperbolic-tangent-based fit to the 1D measurements is extracted and used to define, among other things, the pinch-off, transition, and saturation regions for each gate.…”