Attempt at implementing optimal cutting of a single gate

[garrison]

This is my attempt to implement the optimal cutting of a single two-qubit gate following https://arxiv.org/abs/2312.11638. I have written out the local operations in terms of the operations in https://arxiv.org/abs/2006.11174 and combined like terms, and everything works. However, my new test that 1 <= gamma <= 7 fails, so I think this may be equivalent to the previous code in terms of overhead and therefore not actually optimal 😞. At the moment, I am not sure if it is going to be possible to implement the local channels in the new paper given only single-qubit unitaries and projective measurements (i.e., without projective measurements on ancilla qubits). I will keep thinking about these channels and ask around.

Fixes #531.

[garrison]

There is a revised version of the paper out, at https://arxiv.org/abs/2312.11638v3, which describes the operations in a new appendix. The general operation does indeed require an ancilla qubit on each side of the cut.