Evaluation of synthetic and experimental training data in supervised machine learning applied to charge state detection of quantum dots

Darulová, Jana; Troyer, Matthias; Cassidy, Maja

doi:10.48550/arxiv.2005.08131

Cited by 3 publications

(4 citation statements)

References 29 publications

(46 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The simulation of realistic CSDs requires the consideration of occurring distortions [101], [103], [118]. In the following, we define identified distortion phases, assign collected distortion types to these, and describe their sources, simulation, and the required parameters.…”

Section: Distortions Modelmentioning

confidence: 99%

“…[100] studied the effects of involved quantum parameters on CSDs of a serial triple QD and confirmed their global features by the similarity between transport measurements and CIM-based simulations. To detect charge states, [101] evaluated the prediction accuracy of several machine learning models trained on simulated and experimental data. The simulated data are generated from CIM or taken from the Qflow-lite dataset [102], both improved with five different noise types added.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Simulation of Charge Stability Diagrams for Automated Tuning Solutions (SimCATS)

Hader,

Fleitmann,

Vogelbruch

et al. 2024

Preprint

View full text Add to dashboard Cite

Quantum dots must be tuned precisely to provide a suitable basis for quantum computation. A scalable platform for quantum computing can only be achieved by fully automating the tuning process. One crucial step is to trap the appropriate number of electrons in the quantum dots, typically accomplished by analyzing charge stability diagrams (CSDs). Training and testing automation algorithms require large amounts of data, which can be either measured and manually labeled in an experiment or simulated. This article introduces a new approach to the realistic simulation of such measurements. Our flexible framework enables the simulation of ideal CSD data complemented with appropriate sensor responses and distortions. We suggest using this simulation to benchmark published algorithms. Also, we encourage the extension by custom models and parameter sets to drive the development of robust, technology-independent algorithms.

show abstract

Section: Distortions Modelmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Simulation of Charge Stability Diagrams for Automated Tuning Solutions (SimCATS)

Hader,

Fleitmann,

Vogelbruch

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…Due to the variability inherent in different QD devices, autotuning naturally benefits from a data-driven approach. Several compelling machine learning strategies have recently been introduced for from-scratch QD tuning [10], coarse tuning into charge regimes [11][12][13][14][15], fine tuning couplings between multiple dots [16,17], or performing autonomous measurements [18,19]. These studies have demonstrated * sczischek@uwaterloo.ca robust effectiveness in identifying electronic states and charge configurations, and automating the precise tuning of gates.…”

Section: Introductionmentioning

confidence: 99%

Miniaturizing neural networks for charge state autotuning in quantum dots

Czischek,

Yon,

Genest

et al. 2021

Preprint

View full text Add to dashboard Cite

“…Black box type approaches to coarse tune plunger, barrier and gate voltages to a DQD regime have been developed [10][11][12][13] , allowing tuning devices using direct current measurements faster than a human expert 14 . Furthermore, convolutional neural networks (CNNs) have been trained to determine the charge state of the device at a particular set of gate voltages 12,[15][16][17] . To extract the gradients, Hough transforms have been implemented on a DQD stability diagram in a known charge state 18 , allowing to measure the device in virtual voltage space.…”

Section: Introductionmentioning

confidence: 99%

Automatic virtual voltage extraction of a 2x2 array of quantum dots with machine learning

Oakes,

Duan,

Morton

et al. 2020

Preprint

View full text Add to dashboard Cite

Spin qubits in quantum dots are a compelling platform for fault-tolerant quantum computing due to the potential to fabricate dense two-dimensional arrays with nearest neighbour couplings, a requirement to implement the surface code. However, due to the proximity of the surface gate electrodes cross-coupling capacitances can be substantial, making it difficult to control each quantum dot independently. Increasing the number of quantum dots increases the complexity of the calibration process, which becomes impractical to do heuristically. Inspired by recent demonstrations of industrialgrade silicon quantum dot bilinear arrays, we develop a theoretical framework to cancel the effect of cross-capacitances in a 2×N array of quantum dots, based on the gradients in gate voltage space of different charge transitions that can be measured in multiple two-dimensional charge stability diagrams. To automate the process, we successfully train a neural network to extract the gradients from a Hough transformation of a stability diagram and test the algorithm on simulated and experimental data of a 2×2 quantum dot array.

show abstract

Evaluation of synthetic and experimental training data in supervised machine learning applied to charge state detection of quantum dots

Cited by 3 publications

References 29 publications

Simulation of Charge Stability Diagrams for Automated Tuning Solutions (SimCATS)

Simulation of Charge Stability Diagrams for Automated Tuning Solutions (SimCATS)

Miniaturizing neural networks for charge state autotuning in quantum dots

Automatic virtual voltage extraction of a 2x2 array of quantum dots with machine learning

Contact Info

Product

Resources

About