Description
Correlated noise is a central challenge for scalable quantum error correction: temporal memory and spatial structure can concentrate otherwise rare faults into bursts that evade the assumptions behind standard threshold estimates. However, there remains a gap between microscopic non-Markovian open system dynamics and the stochastic Pauli error models used in stabilizer code simulation and decoding. I will present spatiotemporal Pauli processes (SPPs), an operational framework that closes this gap. An SPP is the multi-time analogue of a Pauli channel: a process-separable Pauli comb, equivalently a joint probability distribution over Pauli fault trajectories in space and time obtained by applying a multi-time Pauli twirl to a general quantum process tensor.
The framework gives a constructive route from physical dynamics to QEC-ready correlated noise models. When the underlying dynamics has a tensor network representation, the Pauli-twirled process inherits a classical tensor network with temporal bond dimension bounded by the effective environment dimension. This makes it possible to diagnose temporal correlations using transfer operators, relate spectral gaps to correlation times, and identify hidden Markov model realisations of physically motivated noise processes.
I will illustrate the framework with surface code memory and stability simulations. First, a tunable temporal "storm" model shows how long-lived memory can degrade distance scaling even at fixed single-round error rates. Second, an SPP derived from a two-dimensional quantum cellular automaton bath produces pseudo-critical error avalanches and a breakdown of conventional distance scaling. These examples demonstrate SPPs as a scalable bridge between open quantum systems theory, correlated noise benchmarking, and fault tolerant quantum computing.
| I am the presenting author | Yes |
|---|