Historically we have avoided the use of 2-qubit gates in constructing preparation and measurement fiducials and this is reflected in the fiducial sets contained within the modelpacks for various gate sets. The restriction to single-qubit gates can significantly increase the number of fiducials required for informational completeness and, while not directly applicable to the modelpacks, for some POVMs two-qubit gates are strictly required to achieve information completeness (native parity readout is one example).
This feature would add an additional kwarg to the fiducial methods for each modelpack that when set would return an alternative set of fiducials (likely a good deal smaller) where 2-qubit gates are allowed. This API will be comparable to the way the lite germ set is handled in the modelpacks.
coreyostrove
commented
1 year ago
Historically we have avoided the use of 2-qubit gates in constructing preparation and measurement fiducials and this is reflected in the fiducial sets contained within the modelpacks for various gate sets. The restriction to single-qubit gates can significantly increase the number of fiducials required for informational completeness and, while not directly applicable to the modelpacks, for some POVMs two-qubit gates are strictly required to achieve information completeness (native parity readout is one example).
This feature would add an additional kwarg to the fiducial methods for each modelpack that when set would return an alternative set of fiducials (likely a good deal smaller) where 2-qubit gates are allowed. This API will be comparable to the way the lite germ set is handled in the modelpacks.