Lie Algebra-Based Quantum Optimal Controls Interpolation
The method combines Lie group theory and neural networks to create a fast way to generate quantum optimal control pulses for arbitrary operations in superconducting qubits. Instead of optimizing every time, it pre-computes a set of representative pulses using Lie algebra sampling and trains a neural network to map target operations directly to these stored pulses. This allows for high-fidelity pulse reconstruction without needing explicit optimization during inference.