A simple method turns readout-error bias in an expectation value into a multiplicative factor that can be measured and removed at the cost of extra sampling. It assumes no specific form of the noise; weak noise is required to avoid excessive sampling overhead.
Abstract
Measurements on current quantum processors are subject to hardware imperfections that lead to readout errors. These errors manifest themselves as a bias in quantum expectation values. Here, we propose a very simple method that forces the bias in the expectation value to appear as a multiplicative factor that can be measured directly and removed at the cost of an increase in the sampling complexity for the observable. The method assumes no specific form of the noise, but only requires that the noise is `weak' to avoid excessive sampling overhead. We provide bounds relating the error in the expectation value to the sample complexity.