Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:2303.00780 · PRX Quantum

Learning Correlated Noise in a 39-Qubit Quantum Processor

Robin Harper, Steven T. Flammia

Published Oct 17, 2023• 35 citations
arXivHTMLPDF

Abstract

Building error-corrected quantum computers relies crucially on measuring and modeling noise on candidate devices.In particular, optimal error correction requires knowing the noise that occurs in the device as it executes the circuits required for error correction.As devices increase in size, we will become more reliant on efficient models of this noise.However, such models must still retain the information required to optimize the algorithms used for error correction.Here, we propose a method of extracting detailed information of the noise in a device running syndrome extraction circuits.We introduce and execute an experiment on a superconducting device using 39 of its qubits in a surface code doing repeated rounds of syndrome extraction but omitting the midcircuit measurement and reset.We show how to extract from the 20 data qubits the information needed to build noise models of various sophistication in the form of graphical models.These models give efficient descriptions of noise in large-scale devices and are designed to illuminate the effectiveness of error correction against correlated noise.Our estimates are furthermore precise: we learn a consistent global distribution where all one-and two-qubit error rates are known to a relative error of 0.1%.By extrapolating our experimentally learned noise models toward lower error rates, we demonstrate that accurate correlated noise models are increasingly important for successfully predicting subthreshold behavior in quantum error-correction experiments.

DOI
10.1103/prxquantum.4.040311

Related papers

Optimal noise estimation from syndrome statistics of quantum codes

Thomas Wagner, Hermann Kampermann, Dagmar Bruß +1 more

Quantum error correction allows to actively correct errors occurring in a quantum computation when the noise is weak enough.To make this error correction competitive information about the specific noise is required.Traditionally, this information is obtained by benchmarking the device before operation.We address the question of what can be learned from only the measurements done during decoding.Such estimation of noise models was proposed for surface codes, exploiting their special structure, and in the limit of low error rates, also for other codes.However, so far it has been unclear under what general conditions noise models can be estimated from the syndrome measurements.In this work, we derive a general condition for identifiability of the error rates.For general stabilizer codes, we prove identifiability under the assumption that the rates are small enough.Without this assumption, we prove a result for perfect codes.Finally, we propose a practical estimation method with linear runtime for concatenated codes.We demonstrate that it outperforms other recently proposed methods and that the estimation is optimal in the sense that it reaches the Cramér-Rao bound.Our method paves the way for practical calibration of error corrected quantum devices during operation.

arXiv:2010.02243•Physical Review Research•Mar 31, 2021•18 citations

Estimating decoding graphs and hypergraphs of memory quantum error-correction experiments

Evangelia Takou, Kenneth R. Brown

Characterizing the error sources of quantum devices is essential for building reliable large-scale quantum architectures and tailoring error correction codes to the noise profile of the devices. Tomography techniques can provide detailed information on the noise or quality of quantum states but are typically both computationally and experimentally intensive with respect to the system's size. For QEC experiments, however, the information captured by a detector error model is sufficient for extracting the success rate of the experiment, as well as some information about the underlying noise. In this work, we estimate Pauli noise on detector error models, including hypergraphs, using only the syndrome statistics. We apply this method to well-known codes such as the repetition, surface, and 2D color codes. Under bare-ancilla syndrome extraction, two-point correlations are enough to reconstruct the detector error model for repetition or surface codes. For color codes or repetition codes under Steane-style syndrome extraction, we show how to extend the estimation method to multi-point correlators and extract the error rates of the hypergraphs. Finally, we find an increase in logical error suppression when we calibrate the decoder to noise fluctuations typically present in experiments.

arXiv:2504.20212•Physical Review A•Nov 7, 2025•16 citations

Exact learning of quantum noise with tensor networks

Nicola Pancotti, Vedika Saravanan, Krysta Svore

Accurate noise models are essential for high-performance quantum error correction, yet characterizing the noise of a quantum device typically requires dedicated experiments. We present a variational framework that learns the noise model directly from quantum-error-correction syndrome and logical-observable data collected during error-corrected memory experiments. The fault-event probabilities are treated as variational parameters and optimized via gradient descent to minimize the binary cross-entropy between the decoder's predictions and experimental logical-observable outcomes. We prove that this objective is principled rather than ad hoc: a sufficiently expressive noise ansatz attains the information-theoretic minimum logical error rate. We instantiate this framework using a tensor-network decoder, which provides exact maximum-likelihood decoding and analytically differentiable gradients with respect to all noise parameters. Using circuit-level data from Google's Sycamore processor, and starting from an uninformed prior, the optimization recovers noise models whose logical error rates agree to within $2\%$ with those of Google's independently characterized detector error model. The mean squared error between the learned and reference noise parameters shows a clear overall decrease throughout training, confirming that the method recovers physically meaningful noise structure, not merely parameters that happen to decode well. We further demonstrate that the optimization can track device drifts in real time via warm-started updates, maintaining near-optimal decoding performance under synthetically evolving noise without the need for re-characterization. The approach is decoder-agnostic in its formulation and naturally extends to correlated noise models.

quant-pharXiv:2609.00169•Sep 2, 2026•3 citations

Scalable Noise Characterization of Syndrome-Extraction Circuits with Averaged Circuit Eigenvalue Sampling

Evan T. Hockings, Andrew C. Doherty, Robin Harper

Characterizing the performance of noisy quantum circuits is central to the production of prototype quantum computers and can enable improved quantum error correction that exploits noise biases identified in a quantum device. We develop a scalable noise-characterization protocol suited to characterizing the syndrome-extraction circuits of quantum error-correcting codes, a key component of fault-tolerant architectures. Our protocol builds upon averaged circuit eigenvalue sampling (ACES), a framework for noise-characterization experiments that simultaneously estimates the Pauli-error probabilities of all gates in a Clifford circuit and captures averaged spatial correlations between gates implemented simultaneously in the layers of the circuit. By rigorously analyzing the performance of noise-characterization experiments in the ACES framework, we derive a figure of merit for their expected performance, allowing us to optimize their experimental design and improve the precision to which we estimate noise given fixed experimental resources. We demonstrate the scalability and performance of our protocol through circuit-level numerical simulations of the entire noise-characterization procedure for the syndrome-extraction circuit of a distance- 25 surface code with over 1000 qubits. Our results indicate that detailed noise-characterization methods are scalable to near-term quantum devices.

arXiv:2404.06545•PRX Quantum•Feb 25, 2025•23 citations

Scalable quantum error correction tailored for a heavy-hex qubit array

Seok-Hyung Lee, Xanda C. Kolesnikow, Jun Zen +7 more

To produce an operable quantum computer that is made with imperfect hardware, we must design and test scalable quantum error correcting codes that are suited for the devices we can build and, in unison, develop decoding strategies that accommodate device-specific noise characteristics. Here, we introduce the \emph{dynamic compass code}, a subsystem code with a novel syndrome extraction cycle, that has a competitive threshold while making efficient use of qubits arranged on a heavy-hex lattice. We use a superconducting qubit array to implement a distance-5 instance of this code, and demonstrate how detailed noise characterisation can boost decoder performance to yield significant improvements in logical error rates. We perform averaged circuit eigenvalue sampling (ACES) to acquire detailed context-dependent error information on all elements of the syndrome extraction process. Furthermore, we leverage soft information produced from measurement devices to augment the decoder with measurement error information and detect leakage errors for exclusion through post-selection. Our noise-informed approach yields up to 38.3\% improvement in the logical error rate of a distance-5 implementation of the dynamic compass code in experiment.

quant-pharXiv:2604.14296•Apr 17, 2026•8 citations

Data-efficient quantum noise modeling via machine learning

Yanjun Ji, Marco Roth, David A. Kreplin +2 more

Maximizing the computational utility of near-term quantum processors requires predictive noise models that inform robust, noise-aware compilation and error mitigation. Conventional models often fail to capture the complex error dynamics of real hardware or require prohibitive characterization overhead. We introduce a data-efficient framework that first constructs a physically motivated, parameterized noise model, and subsequently employs machine-learning-driven Bayesian optimization to identify its parameters. Our approach circumvents costly characterization protocols by estimating algorithm- and hardware-specific error parameters directly from readily available experimental data derived from existing application and benchmark circuit executions. The generality and robustness of the framework are demonstrated across diverse algorithms and superconducting devices, yielding high-fidelity predictions by estimating an independent parameter set tailored to each specific algorithm-hardware context. Crucially, we show that a model calibrated exclusively on small-scale circuits accurately predicts the behavior of larger validation circuits. Our data-efficient approach achieves up to a 65% improvement in model fidelity quantified by the Hellinger distance between predicted and experimental circuit output distributions, compared to standard noise models derived from device properties. This work establishes a practical paradigm for application-aware noise characterization, enabling compilation and error-mitigation strategies tailored to the specific interplay between quantum algorithms and device-specific noise dynamics.

arXiv:2509.12933•Physical Review Applied•Feb 11, 2026•5 citations
Building error-corrected quantum computers relies crucially on measuring and modeling noise on candidate devices.In particular, optimal error correction requires knowing the noise that occurs in the device as it executes the circuits required for error correction.As devices increase in size, we will become more reliant on efficient models of this noise.However, such models must still retain the information required to optimize the algorithms used for error correction.Here, we propose a method of extracting detailed information of the noise in a device running syndrome extraction circuits.We introduce and execute an experiment on a superconducting device using 39 of its qubits in a surface code doing repeated rounds of syndrome extraction but omitting the midcircuit measurement and reset.We show how to extract from the 20 data qubits the information needed to build noise models of various sophistication in the form of graphical models.These models give efficient descriptions of noise in large-scale devices and are designed to illuminate the effectiveness of error correction against correlated noise.Our estimates are furthermore precise: we learn a consistent global distribution where all one-and two-qubit error rates are known to a relative error of 0.1%.By extrapolating our experimentally learned noise models toward lower error rates, we demonstrate that accurate correlated noise models are increasingly important for successfully predicting subthreshold behavior in quantum error-correction experiments.
Quantum error correction allows to actively correct errors occurring in a quantum computation when the noise is weak enough.To make this error correction competitive information about the specific noise is required.Traditionally, this information is obtained by benchmarking the device before operation.We address the question of what can be learned from only the measurements done during decoding.Such estimation of noise models was proposed for surface codes, exploiting their special structure, and in the limit of low error rates, also for other codes.However, so far it has been unclear under what general conditions noise models can be estimated from the syndrome measurements.In this work, we derive a general condition for identifiability of the error rates.For general stabilizer codes, we prove identifiability under the assumption that the rates are small enough.Without this assumption, we prove a result for perfect codes.Finally, we propose a practical estimation method with linear runtime for concatenated codes.We demonstrate that it outperforms other recently proposed methods and that the estimation is optimal in the sense that it reaches the Cramér-Rao bound.Our method paves the way for practical calibration of error corrected quantum devices during operation.
Characterizing the error sources of quantum devices is essential for building reliable large-scale quantum architectures and tailoring error correction codes to the noise profile of the devices. Tomography techniques can provide detailed information on the noise or quality of quantum states but are typically both computationally and experimentally intensive with respect to the system's size. For QEC experiments, however, the information captured by a detector error model is sufficient for extracting the success rate of the experiment, as well as some information about the underlying noise. In this work, we estimate Pauli noise on detector error models, including hypergraphs, using only the syndrome statistics. We apply this method to well-known codes such as the repetition, surface, and 2D color codes. Under bare-ancilla syndrome extraction, two-point correlations are enough to reconstruct the detector error model for repetition or surface codes. For color codes or repetition codes under Steane-style syndrome extraction, we show how to extend the estimation method to multi-point correlators and extract the error rates of the hypergraphs. Finally, we find an increase in logical error suppression when we calibrate the decoder to noise fluctuations typically present in experiments.
Accurate noise models are essential for high-performance quantum error correction, yet characterizing the noise of a quantum device typically requires dedicated experiments. We present a variational framework that learns the noise model directly from quantum-error-correction syndrome and logical-observable data collected during error-corrected memory experiments. The fault-event probabilities are treated as variational parameters and optimized via gradient descent to minimize the binary cross-entropy between the decoder's predictions and experimental logical-observable outcomes. We prove that this objective is principled rather than ad hoc: a sufficiently expressive noise ansatz attains the information-theoretic minimum logical error rate. We instantiate this framework using a tensor-network decoder, which provides exact maximum-likelihood decoding and analytically differentiable gradients with respect to all noise parameters. Using circuit-level data from Google's Sycamore processor, and starting from an uninformed prior, the optimization recovers noise models whose logical error rates agree to within 2%2\%2% with those of Google's independently characterized detector error model. The mean squared error between the learned and reference noise parameters shows a clear overall decrease throughout training, confirming that the method recovers physically meaningful noise structure, not merely parameters that happen to decode well. We further demonstrate that the optimization can track device drifts in real time via warm-started updates, maintaining near-optimal decoding performance under synthetically evolving noise without the need for re-characterization. The approach is decoder-agnostic in its formulation and naturally extends to correlated noise models.
Characterizing the performance of noisy quantum circuits is central to the production of prototype quantum computers and can enable improved quantum error correction that exploits noise biases identified in a quantum device. We develop a scalable noise-characterization protocol suited to characterizing the syndrome-extraction circuits of quantum error-correcting codes, a key component of fault-tolerant architectures. Our protocol builds upon averaged circuit eigenvalue sampling (ACES), a framework for noise-characterization experiments that simultaneously estimates the Pauli-error probabilities of all gates in a Clifford circuit and captures averaged spatial correlations between gates implemented simultaneously in the layers of the circuit. By rigorously analyzing the performance of noise-characterization experiments in the ACES framework, we derive a figure of merit for their expected performance, allowing us to optimize their experimental design and improve the precision to which we estimate noise given fixed experimental resources. We demonstrate the scalability and performance of our protocol through circuit-level numerical simulations of the entire noise-characterization procedure for the syndrome-extraction circuit of a distance- 25 surface code with over 1000 qubits. Our results indicate that detailed noise-characterization methods are scalable to near-term quantum devices.
To produce an operable quantum computer that is made with imperfect hardware, we must design and test scalable quantum error correcting codes that are suited for the devices we can build and, in unison, develop decoding strategies that accommodate device-specific noise characteristics. Here, we introduce the \emph{dynamic compass code}, a subsystem code with a novel syndrome extraction cycle, that has a competitive threshold while making efficient use of qubits arranged on a heavy-hex lattice. We use a superconducting qubit array to implement a distance-5 instance of this code, and demonstrate how detailed noise characterisation can boost decoder performance to yield significant improvements in logical error rates. We perform averaged circuit eigenvalue sampling (ACES) to acquire detailed context-dependent error information on all elements of the syndrome extraction process. Furthermore, we leverage soft information produced from measurement devices to augment the decoder with measurement error information and detect leakage errors for exclusion through post-selection. Our noise-informed approach yields up to 38.3\% improvement in the logical error rate of a distance-5 implementation of the dynamic compass code in experiment.
Maximizing the computational utility of near-term quantum processors requires predictive noise models that inform robust, noise-aware compilation and error mitigation. Conventional models often fail to capture the complex error dynamics of real hardware or require prohibitive characterization overhead. We introduce a data-efficient framework that first constructs a physically motivated, parameterized noise model, and subsequently employs machine-learning-driven Bayesian optimization to identify its parameters. Our approach circumvents costly characterization protocols by estimating algorithm- and hardware-specific error parameters directly from readily available experimental data derived from existing application and benchmark circuit executions. The generality and robustness of the framework are demonstrated across diverse algorithms and superconducting devices, yielding high-fidelity predictions by estimating an independent parameter set tailored to each specific algorithm-hardware context. Crucially, we show that a model calibrated exclusively on small-scale circuits accurately predicts the behavior of larger validation circuits. Our data-efficient approach achieves up to a 65% improvement in model fidelity quantified by the Hellinger distance between predicted and experimental circuit output distributions, compared to standard noise models derived from device properties. This work establishes a practical paradigm for application-aware noise characterization, enabling compilation and error-mitigation strategies tailored to the specific interplay between quantum algorithms and device-specific noise dynamics.