At 400 code mark today with addition of the rotated subsystem surface code. We also have more classical codes now than quantum ones (as it should be) at 204 vs 190. errorcorrectionzoo.org/c/subsy

While the new work uses reinforcement learning, it is based on a theory paper whose supplement quite nicely explains the basics of "infinitesimal" GKP QEC using only a qubit ancilla. arxiv.org/abs/2009.07941 5/5

Show thread

This result should dispel any doubt that it is practically possible to utilize QEC for extending quantum coherence. Since this is a single mode code, scaling the logical lifetime to arbitrary high values still requires concatenation. 4/5

Show thread

Such bosonic platforms are at a technological stage where the overhead brought on by error correction actually helps. Qubit platforms are now approaching break even (plot here is in terms of channel fidelity), and this result pulls bosonic codes somewhat ahead. 3/5

Show thread

Quantum information stored in a GKP code in a microwave cavity has been actively maintained using cycles of QEC whose structure is determined by a reinforcement learner. errorcorrectionzoo.org/c/gkp 2/5

Show thread

QEC has gone beyond break even! Logical error rate of the encoded information is for the first time higher than a bare reference rate. Paper by Vlad Sivak and collaborators @Yale_QI arxiv.org/abs/2211.09116

The Quantum Mastodon

A community of scientist and engineers. Quantum tech (computing, sensing, communication) is close to our hearts!