Calibration of decoders for quantum error correction using multi-agent reinforcement learning

Accurate decoding of quantum error-correcting codes is a crucial ingredient in protecting quantum information from decoherence. It requires characterizing the error channels corrupting the logical quantum state and providing this information as a prior to the decoder, a process known as decoder calibration. In Ref. [1], we introduced a calibration method inspired by multi-agent reinforcement learning, that aims to minimize the logical error rate. It was recently used for calibrating decoders in an experiment that demonstrated below-threshold error correction with the surface code in Ref. [2]. This calibration approach will serve as an important tool for maximizing the performance of both near-term and future error-corrected quantum devices.

[1] V. Sivak et al, Phys. Rev. Lett. 133, 150603 (2024)

[2] Google Quantum AI, arXiv:2408.13687 (2024)