Conditionally rigorous mitigation of multiqubit measurement errors

TL;DR

This paper introduces a conditionally rigorous measurement error mitigation method for quantum computers that is robust against state-preparation errors, improving measurement accuracy in quantum experiments.

Contribution

It develops a measurement error mitigation technique that is insensitive to state-preparation errors, addressing a key limitation of existing methods.

Findings

Effective mitigation of measurement errors on IBM Q superconducting qubits.

Enhanced quantum process tomography with full SPAM error correction.

Demonstrated importance for high-precision quantum measurements.

Abstract

Several techniques have been recently introduced to mitigate errors in near-term quantum computers without the overhead required by quantum error correcting codes. While most of the focus has been on gate errors, measurement errors are significantly larger than gate errors on some platforms. A widely used {\it transition matrix error mitigation} (TMEM) technique uses measured transition probabilities between initial and final classical states to correct subsequently measured data. However from a rigorous perspective, the noisy measurement should be calibrated with perfectly prepared initial states and the presence of any state-preparation error corrupts the resulting mitigation. Here we develop a measurement error mitigation technique, conditionally rigorous TMEM, that is not sensitive to state-preparation errors and thus avoids this limitation. We demonstrate the importance of the technique for high-precision measurement and for quantum foundations experiments by measuring Mermin polynomials on IBM Q superconducting qubits. An extension of the technique allows one to correct for both state-preparation and measurement (SPAM) errors in expectation values as well; we illustrate this by giving a protocol for fully SPAM-corrected quantum process tomography.