Pauli Error Model

qc/noise

A more structured alternative to Depolarizing Noise: instead of randomizing uniformly, apply one specific Pauli operator ($I$, $X$, $Y$, or $Z$) with some probability after each gate.

$$\rho \mapsto p_I\rho +p_{X}X\rho X+p_{Y}Y\rho Y+p_{Z}Z\rho Z,p_I=1-p_X-p_Y-p_Z$$

If $p_X=p_Y=p_Z=\lambda /4$ (uniform), this reduces exactly to the depolarizing model. The key extra power here is directionality — noise can be stronger along one axis than another, which is physically realistic (real hardware noise is rarely isotropic).

This directionality is exactly what Bit-flip and Phase-flip Sensitivity exploits: by choosing which basis you measure in, you can selectively reveal or hide different Pauli components of the error.

Error	Bit-flip sensitive (Z-basis)	Phase-flip sensitive (X-basis)
$I$	No	No
$X$	Yes	No
$Y$	Yes	Yes
$Z$	No	Yes

Quick mnemonic: bit-flip sensitivity ↔ “does this error contain an $X$?”; phase-flip sensitivity ↔ “does this error contain a $Z$?” — reading straight off the identity $Y=iXZ$ explains why $Y$ trips both.

Related

• Bit-flip and Phase-flip Sensitivity
• Depolarizing Noise
• Density Matrix
• Thermal Relaxation — sibling toy noise model, but time-dependent rather than per-gate
• Pauli Operators
• Coherent vs Incoherent Gate Errors — this model describes incoherent (stochastic Pauli) error specifically
• Physical Decoherence Mechanisms

Self-Check

• What extra realism does the Pauli error model add over depolarizing noise?
• Using the $Y=iXZ$ mnemonic, could you explain why a $Y$ error trips both bit-flip and phase-flip sensitivity?
• Under what condition does the Pauli error model reduce exactly to the depolarizing model?