Determine the source of asymmetric Pauli noise

Identify the physical or implementation-level origin of the strong asymmetry in the Pauli noise channel observed in the IBM Heron r2 quantum-computer simulations, particularly why measurements of the single-qubit observable \(\langle \hat{Z}_n\rangle\) are approximately three times noisier than measurements of \(\langle \hat{X}_n\rangle\) and \(\langle \hat{Z}_n\hat{Z}_{n+1}\rangle\), and determine whether incomplete twirling of the non-Clifford \(R_{ZZ}\) gate is responsible.

Background

The quantum simulations of inelastic scattering in one-dimensional Ising field theory use Pauli twirling to convert hardware noise into an effective stochastic Pauli channel. The authors observe that this channel is substantially asymmetric: single-qubit $Z$-measurements are approximately three times noisier than corresponding $X$- and nearest-neighbor $ZZ$-measurements. This asymmetry complicates error mitigation because the reliability of the measured observables depends strongly on their Pauli type.

The authors do not determine the origin of this behavior. They suggest that incomplete twirling of the non-Clifford $R_{ZZ}$ gate may be responsible, but characterize this only as a possibility. Resolving the source would clarify the device-noise model and could improve error mitigation for large-light-cone quantum simulations.