CutReg: A loss regularizer for enhancing the scalability of QML via adaptive circuit cutting

2025-06-17

AI Summary

Quantum machine learning (QML) on NISQ devices is hindered by shallow circuit depth, limited qubit connectivity, and challenges in verifying quantum advantage or diagnosing barren plateaus; while circuit cutting enables execution on small-scale hardware, it incurs substantial sampling overhead. Method: We propose a sampling-overhead-aware regularization framework that explicitly models the variance of expectation-value estimates—introduced by circuit cutting—as a differentiable regularizer, enabling end-to-end joint optimization of both cutting strategies and QML parameters. Our approach integrates circuit cutting theory, stochastic measurement estimation, and gradient backpropagation to support differentiable training of parameterized quantum circuits. Contribution/Results: Experiments demonstrate that our method significantly reduces equivalent sampling complexity while preserving model accuracy, thereby enhancing the scalability and training efficiency of QML on NISQ hardware and establishing a new paradigm for empirical quantum advantage investigation.

Technology Category

• Machine Learning: Quantum Machine Learning
• Search and Optimization: Sampling/Simulation-based Search
• Natural Language Processing: Learning & Optimization for NLP

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Security and Privacy: Security and privacy of machine learning and AI applications
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Whether QML can offer a transformative advantage remains an open question. The severe constraints of NISQ hardware, particularly in circuit depth and connectivity, hinder both the validation of quantum advantage and the empirical investigation of major obstacles like barren plateaus. Circuit cutting techniques have emerged as a strategy to execute larger quantum circuits on smaller, less connected hardware by dividing them into subcircuits. However, this partitioning increases the number of samples needed to estimate the expectation value accurately through classical post-processing compared to estimating it directly from the full circuit. This work introduces a novel regularization term into the QML optimization process, directly penalizing the overhead associated with sampling. We demonstrate that this approach enables the optimizer to balance the advantages of gate cutting against the optimization of the typical ML cost function. Specifically, it navigates the trade-off between minimizing the cutting overhead and maintaining the overall accuracy of the QML model, paving the way to study larger complex problems in pursuit of quantum advantage.

Problem

• Enhancing QML scalability via adaptive circuit cutting
• Reducing sampling overhead in partitioned quantum circuits
• Balancing cutting benefits with ML cost optimization

Innovation

• Adaptive circuit cutting for QML scalability
• Regularization term to penalize sampling overhead
• Balancing gate cutting and ML cost optimization