In certain situations, it is convenient to define a unitary infidelity loss that takes into account the ability of the hardware to easily handle virtual Z-gates. A loss can be written down for this in the form of
Here $\theta$ is a vector of free Z-rotation angles, that are also decision variables. An issue in NamedTrajectories.jl will provide the functionality to store "global" parameters -- parameters that don't depend on time -- to be used as decision variables in the optimizer.
Completing this issue requires:
[ ] implementing functionality to handle global parameters stored in a NamedTrajectory
[ ] implementing the loss defined above with derivatives (either autodiff or tested analytic derivatives)
[ ] documentation for the use of this objective
[ ] a problem template for setting up these types of problems
Feature Description
In certain situations, it is convenient to define a unitary infidelity loss that takes into account the ability of the hardware to easily handle virtual Z-gates. A loss can be written down for this in the form of
$$ \ell(U, \theta) = \mathcal{F}\left( \left(\bigotimes\alpha Z(\theta\alpha) \right) U, U_{\text{goal}} \right). $$
Here $\theta$ is a vector of free Z-rotation angles, that are also decision variables. An issue in NamedTrajectories.jl will provide the functionality to store "global" parameters -- parameters that don't depend on time -- to be used as decision variables in the optimizer.
Completing this issue requires:
NamedTrajectory
Importance
3
What does this feature affect?
Other information
No response