Deep Learning Approaches to Quantum Error Mitigation

TL;DR

This paper explores deep learning models, especially attention-based architectures, for quantum error mitigation, demonstrating their effectiveness on real quantum hardware and outperforming traditional methods.

Contribution

It systematically compares neural network architectures for quantum error mitigation and introduces attention-based models as the most effective approach.

Findings

Attention-based models outperform other architectures.

Models generalize well across similar quantum devices.

Deep learning methods improve error mitigation over baseline techniques.

Abstract

We present a systematic investigation of deep learning methods applied to quantum error mitigation of noisy output probability distributions from measured quantum circuits. We compare different architectures, from fully connected neural networks to transformers, and we test different design/training modalities, identifying sequence-to-sequence, attention-based models as the most effective on our datasets. These models consistently produce mitigated distributions that are closer to the ideal outputs when tested on both simulated and real device data obtained from IBM superconducting quantum processing units (QPU) up to five qubits. Across several different circuit depths, our approach outperforms other baseline error mitigation techniques. We perform a series of ablation studies to examine: how different input features (circuit, device properties, noisy output statistics) affect performance; cross-dataset generalization across circuit families; and transfer learning to a different IBM QPU. We observe that generalization performance across similar devices with the same architecture works effectively, without needing to fully retrain models.