Benchmarking machine learning models for multi-class state recognition in double quantum dot data

TL;DR

This study benchmarks four machine learning architectures for multi-class state recognition in double quantum dot charge-stability diagrams, highlighting CNNs with min-max normalization as the most practical solution.

Contribution

It provides a comprehensive comparison of ML models for quantum dot state recognition, emphasizing CNNs' effectiveness and the impact of normalization schemes.

Findings

U-Nets and ViTs perform best on synthetic data but not on experimental data.

CNNs achieve a good balance of accuracy and efficiency on experimental data.

Normalization affects training stability and accuracy, with min-max favoring accuracy and z-score favoring stability.

Abstract

Semiconductor quantum dots (QDs) are a leading platform for scalable quantum processors. However, scaling to large arrays requires reliable, automated tuning strategies for devices' bootstrapping, calibration, and operation, with many tuning aspects depending on accurately identifying QD device states from charge-stability diagrams (CSDs). In this work, we present a comprehensive benchmarking study of four modern machine learning (ML) architectures for multi-class state recognition in double-QD CSDs. We evaluate their performance across different data budgets and normalization schemes using both synthetic and experimental data. We find that the more resource-intensive models -- U-Nets and visual transformers (ViTs) -- achieve the highest MSE score (defined as $1-\mathrm{MSE}$) on synthetic data (over $0.98$) but fail to generalize to experimental data. MDNs are the most computationally efficient and exhibit highly stable training, but with substantially lower peak performance. CNNs offer the most favorable trade-off on experimental CSDs, achieving strong accuracy with two orders of magnitude fewer parameters than the U-Nets and ViTs. Normalization plays a nontrivial role: min-max scaling generally yields higher MSE scores but less stable convergence, whereas z-score normalization produces more predictable training dynamics but at reduced accuracy for most models. Overall, our study shows that CNNs with min-max normalization are a practical approach for QD CSDs.