Experimental performance of survival-probability fidelity estimation under SPAM errors

Establish experimentally the relative performance of the survival-probability-based adjoint-cycle fidelity estimator and direct fidelity estimation under readout noise and other state-preparation-and-measurement errors.

Background

The survival-probability method uses repeated return-probability measurements and may reduce some sensitivities to state-preparation-and-measurement errors compared with direct fidelity estimation. However, the paper’s simulations use exact moments of the positive adjoint cycle and do not include finite-shot fluctuations or explicit SPAM errors.

Consequently, the comparative robustness of the two protocols under realistic readout noise and broader SPAM imperfections has not been resolved experimentally.

References

Its relative performance under readout noise and other SPAM errors remains to be established experimentally.

— Practical Fidelity Estimation for Noisy Quantum Devices Using Direct and Survival-Probability-Based Protocols  (2610.21606 - Aman et al., 15 Oct 2026) in Section Results and Discussion, subsection “Comparison of the two protocols”