arXiv:2610.19034
Logical gate set tomography
Piper C. Wysocki, Juan Gonzalez De Mendoza, Karl Mayer, Natalie C. Brown, Ciaran Ryan-Anderson, Jalan A. Ziyad, Jordan H. Hines, Daniel G. Hothem, Corey I. Ostrove, Timothy J. Proctor, Kevin C. Young, Robin J. Blume-Kohout, Charles H. Baldwin, Kenneth M. Rudinger
Abstract
Quantum error correction (QEC) can strongly suppress noise, but predicting whether a computation succeeds requires understanding not only how much effective noise persists after QEC, but what form that noise takes. To thoroughly characterize this noise, we introduce logical gate set tomography (``logical GST''). Logical GST produces detailed representations of noisy logical operations that can be used to derive higher-level metrics, such as gate fidelity and diamond distance, and to test for non-Markovian and coherent logical errors. We use logical GST to obtain the first full, self-consistent characterization of a universal single-qubit logical gate set, implemented in the seven-qubit Steane code using Quantinuum System Model H1, a 20-qubit trapped-ion quantum processor. In the experimental data from H1, logical GST identified predominantly logical Pauli-stochastic errors, found no evidence of logical non-Markovianity, and detected weak potential signatures of coherent logical error. By contrast, higher-shot datasets generated with Quantinuum's detailed system emulator, identify mechanisms for weak but statistically significant logical non-Markovianity and strong evidence of coherent logical error. Finally, we use logical GST to perform error attribution with the system emulator, revealing how specific physical error sources shape logical error mechanisms.
AI Summary
Logical error rates alone do not determine how errors accumulate in fault-tolerant circuits. This paper adapts gate set tomography (GST) to encoded operations, estimating self-consistent logical channels while testing whether fixed Markovian channels and Pauli-stochastic noise adequately describe the data. The experimental demonstration characterizes logical preparation, measurement, idle, $H_L$, $S_L$, and magic-state-injected $T_L$ operations in the $\llbracket7,1,3\rrbracket$ Steane code on Quantinuum H1-1, using active correction. A compressed design uses 158 circuits with approximately 50 retained shots each, reaching 38 QEC cycles. Maximum-likelihood CPTP fits give entanglement infidelities of $(0.7\pm0.4)\%$, $(0.2\pm0.4)\%$, and $(1.2\pm1.5)\%$ for $H_L$, $S_L$, and $T_L$, respectively, with 95% intervals. Hardware data show no significant non-Markovianity or rejection of Pauli-stochastic noise, although individual coherent axis-misalignment parameters modestly improve fits. Emulator datasets use 931 circuits and 14,100 shots per circuit, independently toggling leakage and coherent memory errors. Likelihood tests detect strong departures from Pauli-stochastic noise when memory errors are enabled and weaker non-Markovianity, requiring unmodeled error below 1% of per-operation diamond-distance error. Logical over-rotations near $0.0015\pi$ are resolved for $S_L$, idle, and $T_L$; an independent logical Ramsey experiment corroborates the $S_L$ estimate. Automated model selection removes at least 22 of 43 observable parameters. Limitations include low hardware statistical power, emulator dependence, and implementation-specific distance-three behavior. The $T_L$ reconstruction conditions on post-selection acceptance. Mechanisms underlying emergent non-Markovianity remain unresolved, and Pauli-frame tracking could change the observed noise.
the authors used AI-based tools to assist in generating code for analysis and visualization and to support manuscript outlining, drafting, and prose development.
Error found: The abstract's full-characterization claim omits that the reconstructed T_L operation is conditioned on successful injection and excludes rejected outcomes. The abstract overstates the identification of emergent logical non-Markovianity mechanisms relative to the appendix's explicitly tentative conclusions. The evidence-ratio equation has the wrong denominator sign under the paper's definition of kappa, although its reported values and interpretation use a positive ratio. The appendix's proposed explanation of residual T_L coherent error yields a symmetric stochastic-channel generator with no first-order Hamiltonian component, so its commutator rewriting does not establish the claimed coherent mechanism.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.