Closed-Loop Optimized Control Strategies for Adiabatic Quantum Computation in the Presence of Noise

We extend recent work focused on closed-loop optimization for adiabatic quantum computation (AQC) control schedules and assess the robustness of a number of closed-loop noise mitigation strategies in the open system setting. First, we investigate the robustness of the previously developed approach in the open system setting and characterize optimal performance regimes with respect to noise parameters. Second, we modify the approach to optimize amplitudes of terms in the AQC Hamiltonian, such as the local bias and tunneling strengths of the initial Hamiltonian. We find that the latter approach can be considerably more robust than schedule optimization for random Ising problem Hamiltonians. Lastly, the technique is extended to stabilizer subsystem code encoded AQC evolution to optimize the static and time-dependent amplitudes of energy penalties, where considerable improvements over uniform static energy penalties is observed for random Ising chain problems.