Optimizing continuous-time quantum error correction for arbitrary noise

We present a protocol using machine learning (ML) to simultaneously optimize the quantum error-correcting code space and the corresponding recovery map in the framework of continuous-time quantum error correction. Given a Hilbert space and a noise process – potentially correlated across both space and time – the protocol identifies the optimal recovery strategy, measured by the average logical state fidelity. This approach enables the discovery of recovery schemes tailored to arbitrary device-level noise.

💡 Research Summary

The paper introduces a machine‑learning‑driven framework for jointly optimizing the code subspace and the recovery map in continuous‑time quantum error correction (CT‑QEC) under arbitrary noise. Traditional stabilizer codes assume Pauli‑type, Markovian errors and rely on discrete, fast syndrome measurements followed by unitary corrections. Real quantum hardware, however, experiences non‑Markovian dynamics, leakage, and non‑Pauli errors, making standard channel‑adapted schemes sub‑optimal, especially when measurements and corrections must act simultaneously in the continuous‑time limit.

The authors formulate the dynamics as ρ̇ = (𝔇_N + 𝔇_R)(ρ), where 𝔇_N describes the noise and 𝔇_R the recovery. The performance metric is the average logical fidelity F̄ = E_noise