Author(s): Remmy Zen, Jan Olle, Luis Colmenarez, Matteo Puviani, Markus Müller, and Florian Marquardt Using reinforcement learning to design fault-tolerant quantum circuits leads to efficient logical state preparation schemes with fewer gates and flag qubits than human-designed methods, advancing quantum error correction. [Phys. Rev. X 15, 041012] Published Wed Oct 22, 2025