False Positives Raised by Quantum Readout Error Mitigation

Quantum readout error mitigation is essential for noisy intermediate-scale quantum devices to achieve reliable data. The conventional approaches, conflating initialization errors with measurement errors, not only suppress the influence of measurement errors, but also strengthen that of initialization errors, which is a systematic bias grows exponentially with the qubit number. Here, we have proved that this effect causes severe fidelity overestimation for all stabilizer states and might lead to false positives in large-scale entangled state characterization. Similarly, the results from algorithms like the variational quantum eigensolver and time evolution also deviate negatively, and cover up other errors in the quantum circuit. These findings highlight the critical need for rigorous benchmarking and careful management of initialization errors. Consequently, we establish an upper bound for the tolerable initialization error rate to ensure effective error mitigation at a given system scale.

💡 Research Summary

In this work the authors expose a hidden systematic bias in the widely used quantum readout error mitigation (QREM) technique that stems from neglecting state‑preparation (initialization) errors. Conventional QREM assumes that the measured probability distribution P_noisy is related to the ideal distribution P_ideal by a linear map P_noisy = M P_ideal, where M is the measurement‑error matrix obtained by preparing all computational basis states and measuring the outcomes. The mitigation step then applies the inverse of M (or a regularized version) to recover P_ideal. However, when each qubit suffers an initialization error with probability q_i, the effective map becomes Λ M, where Λ incorporates the preparation errors. Even if q_i ≪ measurement‑error rates, the product Λ M deviates from M, and the deviation compounds exponentially with the number of qubits n.

The authors first develop a rigorous analytical model for stabilizer states. For an n‑qubit stabilizer state |φ⟩, the noisy density matrix can be written as ρ_noisy = ½