Robust two-qubit gate with reinforcement learning and dropout

Abstract

In the realm of quantum control, reinforcement learning, a prominent branch of machine learning, emerges as a competitive candidate for computer-assisted optimal experiment design. This paper investigates the extent to which guidance from human experts is necessary for effectively implementing reinforcement learning in the design of quantum control protocols. Specifically, our focus lies on engineering a robust two-qubit gate, utilizing a combination of analytical solutions as prior knowledge and techniques from computer science. Through thorough benchmarking of various models, we identify dropout—a widely used method for mitigating overfitting in machine learning—as a particularly robust approach. Our findings demonstrate the potential of integrating computer science concepts to propel the development of advanced quantum technologies.