Improving stabilizer approximation with quantum strategy

We introduce a quantum strategy from nonlocal games to improve the stabilizer approximation we proposed previously. The resulting approach turns out to be a qubit-by-qubit gauging procedure for standard stabilizers, which could involve discrete or continuous gauge parameters. We take examples from many-body physics and quantum chemistry to show such a procedure leads to an improvement of the performance.