Quest: Graph transformer for quantum circuit reliability estimation

This paper presents a case study of using machine learning (ML) for quantum system research, specifically for predicting the impact of quantum noise on circuit fidelity. Inspired by the natural graph representation of quantum circuits, the authors propose to leverage a graph transformer model to predict the noisy circuit fidelity. The proposed method involves collecting a large dataset with various quantum circuits and their fidelity on noisy simulators and real machines, embedding each circuit into a graph with gate and noise properties as node features, and adopting a graph transformer to predict the fidelity. Evaluated on 5 thousand random and algorithm circuits, the graph transformer predictor can provide accurate fidelity estimation with RMSE error 0.04 and outperform a simple neural network-based model by 0.02 on average. PDF