Visual explanations of machine learning models to estimate charge states in quantum dots

A group of researchers has successfully demonstrated automatic charge state recognition in quantum dot devices using machine learning techniques, representing a significant step toward automating the preparation and tuning ...
Details of the research were published in the journal APL Machine Learning on April 15, 2024.

Semiconductor use to create quantum bits. These materials are common in traditional electronics, making them integrable with conventional semiconductor technology. This compatibility is why scientists consider them strong candidates for future qubits in the quest to realize quantum computers.
In semiconductor spin qubits, the spin state of an electron confined in a quantum dot serves as the fundamental unit of data, or the qubit. Forming these qubit states requires tuning numerous parameters, such as gate voltage, something performed by human experts.

However, as the number of qubits grows, tuning becomes more complex due to the excessive number of parameters. When it comes to realizing large-scale computers, this becomes problematic.

"To overcome this, we developed a means of automating the estimation of charge states in double quantum dots, crucial for creating spin qubits where each quantum dot houses one electron," says Tomohiro Otsuka, an associate professor at Tohoku University's Advanced Institute for Materials Research (WPI-AIMR).

(a) The flow to train the estimator. Training data for the CNN was prepared by simulation using the CI model. The researchers simplified the data with pre-processing and then trained the CNN. (b) The flow to estimate the charge state in the experimental data. The researchers also simplified the data with pre-processing and then inputted the trained estimator to estimate the charge state. Credit: APL Machine Learning (2024). DOI: 10.1063/5.0193621
Read more

Figure visualizing the estimator's decision basis in regions where the charge state estimation was correct, using Grad-CAM. Pixels corresponding to the charge transition lines are prominently highlighted. Credit: APL Machine Learning (2024). DOI: 10.1063/5.0193621
Read more

The top figure visualizes the estimator's decision basis in regions where the charge state estimation was erroneous. Pixels where noise coincidentally connected are prominently highlighted, suggesting a possible misidentification as charge transition lines. The bottom figure shows the charge state estimation results for experimental data using the improved estimator. The color on the experimental data indicates the estimated results. The estimator achieved sufficient accuracy. Credit: APL Machine Learning (2024). DOI: 10.1063/5.0193621
Read more