Benchmarking error mitigation techniques on noisy quantum processors

Several error mitigation techniques have been proposed in the recent years to compensate for the errors of noisy, intermediate-scale quantum (NISQ) devices before full-fledged, yet costly error correction methods can be implemented. Here, we present an implementation of a combination of these techniques, and a detailed benchmark based both on realistic noisy simulations and actual computations on current quantum architectures such as, for instance, superconducting transmon processors. We examine the link between the accuracy of the knowledge of the quantum hardware (via tomography), the accuracy on the final observable, and the overhead of mitigation. We also discuss the relevance of these methods for reaching high enough accuracies in applications such as quantum chemistry.