Reinforcement learning (RL) can be leveraged to improve the search for viable quantum circuit architectures, but current RL approaches face significant challenges in this domain.
The authors propose a standard cell approach to efficiently design and compile large-scale quantum circuits, particularly for neutral atom quantum computers. Their method leverages the regular structure of quantum circuits and qubit layouts to enable fast, scalable, and resource-efficient routing and compilation.
This paper introduces a new optimization model and symmetry-breaking constraints to efficiently design minimum-cost quantum circuits using Multiple-Control Toffoli (MCT) gates.
While random number-conserving quantum circuits are useful for various quantum information processing tasks, their finite moments cannot be distinguished from those of the Haar ensemble on the entire group of number-conserving unitaries, except for very high-order moments.