Discovering autonomous quantum error correction via deep reinforcement learning
Researchers have employed deep reinforcement learning with curriculum learning to discover novel autonomous quantum error correction (AQEC) codes. This method aims to overcome the limitations of traditional quantum error correction by using engineered dissipation and drives. The developed agent successfully identified optimal codewords, specifically Fock states \ket{4} and \ket{7}, which demonstrate state-of-the-art performance against single-photon and double-photon losses, and are robust to phase and amplitude damping noise. AI
IMPACT This research demonstrates the potential of AI in discovering complex quantum error correction codes, potentially accelerating the development of fault-tolerant quantum computing.