Optimal control of a superconducting bosonic processor in the presence of noise

Research Project, 2026 – 2030

Continuous-variable quantum computing with bosonic modes in superconducting circuits offers a powerful pathway toward fault-tolerant quantum information processing. Two key experiments at Chalmers demonstrated universal control of bosonic modes: one using a 3D cavity resonator coupled to a qubit, the other a planar resonator terminated by SNAIL. Despite these advances, progress is limited by decoherence processes, including cavity photon loss, qubit relaxation, and dephasing. At present, there is no general analytical framework to quantify how decoherence impacts the fidelity of bosonic operations, making it hard to tell whether errors stem from decoherence or can be reduced with better control. The first goal is to develop such a framework to quantify fidelity loss from decoherence. I will then apply it to experimental data to identify how much improvement can come from better pulse control. The second goal is to develop a robust optimization technique, combining GRAPE and reinforcement learning, tailored to VeriQuB, the multi-mode bosonic platform under development at Chalmers within an EIC-funded project. To enable efficient experimental feedback, I will incorporate my recent method for selective quantum state tomography using only two measurements. I will lead the theory work and collaborate with Chalmers´ experimental teams. With the VeriQuB platform underway, this is a timely opportunity to provide the tools needed to ensure it reaches its full potential.