arxiv: 2603.29439 · v4 · submitted 2026-03-31 · 🪐 quant-ph

Recognition: 2 theorem links

· Lean Theorem

PAEMS: Precise and Adaptive Error Model for Superconducting Quantum Processors

Songhuan He , Yifei Cui , Bo Liu , Kai Guo , Cheng Wang

Authors on Pith no claims yet

Pith reviewed 2026-05-14 00:03 UTC · model grok-4.3

classification 🪐 quant-ph

keywords quantum error modelingsuperconducting qubitsquantum error correctionleakage propagationrepetition codeserror correlationadaptive error model

0 comments

The pith

PAEMS uses qubit-wise separation and leakage propagation optimized on repetition code data to model superconducting qubit errors with far lower correlations than prior models.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces PAEMS, a new error model designed to overcome the limitations of simple depolarizing models and computationally heavy density matrix approaches for superconducting quantum processors. It separates errors at the individual qubit level while tracking leakage effects that spread across space and time, then tunes the entire model end-to-end against data from repetition code experiments. This produces synthetic error data that matches real hardware behavior much more closely, cutting timelike correlations by 19.5 times, spacelike by 9.3 times, and spacetime by 5.2 times on IBM devices while beating Google's SI1000 model by 58-73 percent across several platforms. A sympathetic reader would care because accurate error models let quantum error correction decoders work reliably without requiring impractically large experimental datasets.

Core claim

PAEMS introduces a qubit-wise separation framework incorporating leakage propagation to capture error evolvements crossing spatial and temporal domains. Utilizing repetition-code experiment datasets, it identifies intrinsic qubit errors through an end-to-end optimization pipeline, achieving 19.5×, 9.3×, and 5.2× reductions in timelike, spacelike, and spacetime error correlations on IBM QPUs while outperforming the accuracy of Google's SI1000 model by 58-73% on IBM's Brisbane, Sherbrooke, and Torino as well as China Mobile's Wuyue and China Telecom's Tianyan.

What carries the argument

The qubit-wise separation framework with leakage propagation, which models each qubit's error processes independently while propagating leakage effects across neighboring qubits and time steps, then fits all parameters via end-to-end optimization on repetition code data.

If this is right

QEC decoders trained on PAEMS-generated data achieve lower logical error rates because the modeled correlations more closely match real hardware.
The same optimization pipeline can be re-run on new devices or calibrations to produce platform-specific error models without redesigning the framework.
Leakage propagation tracking enables more accurate simulation of error spread in larger qubit arrays used for surface codes or other QEC schemes.
Reduced reliance on massive experimental datasets lowers the barrier to testing and refining QEC protocols on current superconducting hardware.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The approach may extend to other qubit technologies if suitable calibration experiments can be substituted for repetition codes.
Future decoders could incorporate the PAEMS parameters directly as priors rather than treating them as fixed synthetic inputs.
If the model generalizes across circuit depths, it could reduce the cost of full quantum circuit simulation for error analysis.

Load-bearing premise

Repetition code experiment datasets, when fed into the end-to-end optimization, correctly isolate intrinsic qubit errors rather than fitting to hardware-specific noise patterns or post-selection effects.

What would settle it

Apply the trained PAEMS model to predict error rates and correlations on a fresh set of non-repetition-code circuits executed on the same hardware and check whether measured outcomes match the predictions within the reported accuracy margins.

Figures

Figures reproduced from arXiv: 2603.29439 by Bo Liu, Cheng Wang, Kai Guo, Songhuan He, Yifei Cui.

**Figure 2.** Figure 2: FIG. 2. Repetition code experiment with 21 qubits and 30 detection rounds. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 3.** Figure 3: FIG. 3. Cross-platform adaptivity with single-round repetition code. [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

**Figure 4.** Figure 4: FIG. 4. Framework of PAEMS qubit error model. [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗

**Figure 5.** Figure 5: d∼f compare the qubit calibration of Torino QPU with the optimized parameters by PAEMS. A divergence of qubit parameters with a factor of 3 to 5 is observed, which is notable in the relaxation and dephasing times T1 and T2 of qubit 9, 10, 12, and 13. For example, qubit 9 shows a parameter variation from T1 = 47.9 µs and T2 = 42.3 µs to T1 = 16.7 µs and T2 = 8.1 µs after the optimization. In addition, Fig… view at source ↗

read the original abstract

Superconducting quantum processor units (QPUs) are incapable of producing massive datasets for quantum error correction (QEC) because of hardware limitations. Thus, QEC decoders heavily depend on synthetic data from qubit error models. Classic depolarizing error models with polynomial complexity present limited accuracy. Coherent density matrix methods suffer from exponential complexity $\propto O(4^n)$ where $n$ represents the number of qubits. This paper introduces PAEMS: a precise and adaptive qubit error model. Its qubit-wise separation framework, incorporating leakage propagation, captures error evolvements crossing spatial and temporal domains. Utilizing repetition-code experiment datasets, PAEMS effectively identifies the intrinsic qubit errors through an end-to-end optimization pipeline. Experiments on IBM's QPUs have demonstrated a 19.5$\times$, 9.3$\times$, and 5.2$\times$ reduction in timelike, spacelike, and spacetime error correlation, respectively, surpassing all of the previous works. It also outperforms the accuracy of Google's SI1000 error model by 58$\sim$73\% on multiple quantum platforms, including IBM's Brisbane, Sherbrooke, and Torino, as well as China Mobile's Wuyue and China Telecom's Tianyan.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

PAEMS claims sharp drops in error correlations via qubit-wise fitting on repetition-code data, but the same datasets drive both the model and the reported gains, leaving generalization unclear.

read the letter

Colleague, the paper's main move is a qubit-wise error model that tracks leakage propagation across space and time, then fits parameters end-to-end to repetition-code runs on IBM and other QPUs. It reports 19.5x, 9.3x, and 5.2x reductions in timelike, spacelike, and spacetime correlations, plus 58-73% better accuracy than Google's SI1000 model across several platforms. That is the concrete claim worth checking. The framework itself is straightforward: separate per-qubit rates instead of a global depolarizing channel, add explicit leakage terms, and optimize to match observed syndromes. This is more scalable than full density-matrix simulation and more detailed than basic polynomial models, which is useful for generating training data for decoders. The authors do test on multiple machines (Brisbane, Sherbrooke, Torino, plus two Chinese platforms), which is better than single-device results. The soft spot is the evaluation loop. Parameters come from optimizing on the repetition-code datasets, and the correlation reductions are measured on the same or closely related data. Without explicit held-out circuit families, cross-validation splits, or tests on unrelated gate sequences, it is hard to separate real error isolation from fitting to device-specific patterns like crosstalk or calibration drift. The abstract does not show ablations that isolate the leakage terms or the qubit-wise split from the optimization itself. If those details are in the full methods and the numbers survive independent checks, the practical gain for decoder training could be real. This is for people who build or benchmark QEC decoders and need synthetic noise that is cheap to sample but closer to hardware than depolarizing noise. A reader who already works with repetition-code data or SI1000 baselines will get the most out of it. The work is coherent on its own terms and engages the right literature, so it deserves a serious referee to examine the pipeline and ask for generalization tests. I would send it to review rather than desk reject.

Referee Report

2 major / 2 minor

Summary. The paper introduces PAEMS, a qubit error model for superconducting QPUs that employs a qubit-wise separation framework incorporating leakage propagation to capture spatial and temporal error correlations. Parameters are extracted via an end-to-end optimization pipeline applied to repetition-code experiment datasets from IBM QPUs. The work claims 19.5×, 9.3×, and 5.2× reductions in timelike, spacelike, and spacetime error correlations respectively, along with 58–73% accuracy gains over Google's SI1000 model on IBM Brisbane/Sherbrooke/Torino and other platforms including China Mobile Wuyue and China Telecom Tianyan.

Significance. If the model isolates intrinsic errors that generalize beyond the fitting datasets, it would provide a practical advance for generating accurate synthetic data for QEC decoders, addressing the accuracy limits of depolarizing models and the scaling issues of full density-matrix simulations. The reported correlation reductions are large enough to matter for near-term repetition-code and surface-code experiments.

major comments (2)

[Abstract and §3] Abstract and §3 (optimization pipeline): The manuscript states that parameters are obtained via end-to-end optimization on the same repetition-code datasets used to compute the reported correlation reductions. No description is given of the loss function, regularization, hyperparameter search, or any train/validation split. This makes it impossible to determine whether the 19.5×/9.3×/5.2× improvements reflect genuine isolation of intrinsic errors or device-specific noise signatures present in the IBM runs.
[§4] §4 (results and comparisons): The accuracy gains versus SI1000 (58–73%) and the cross-platform claims are presented without explicit held-out circuit families, ablation of the qubit-wise separation component, or transfer tests on independent datasets. The absence of these controls leaves open the possibility that the reported improvements are partly due to post-hoc fitting rather than a generalizable model.

minor comments (2)

[Abstract and §4] The abstract lists five platforms but the main text provides limited quantitative detail on the China Mobile and China Telecom devices; expanding Table 2 or adding a supplementary table with per-platform metrics would improve clarity.
[§2] Notation for the leakage propagation terms is introduced without an explicit equation reference in the early sections; adding a numbered equation for the leakage operator would aid readability.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback. We address the two major comments point by point below and will revise the manuscript accordingly to improve clarity and strengthen the evaluation.

read point-by-point responses

Referee: [Abstract and §3] Abstract and §3 (optimization pipeline): The manuscript states that parameters are obtained via end-to-end optimization on the same repetition-code datasets used to compute the reported correlation reductions. No description is given of the loss function, regularization, hyperparameter search, or any train/validation split. This makes it impossible to determine whether the 19.5×/9.3×/5.2× improvements reflect genuine isolation of intrinsic errors or device-specific noise signatures present in the IBM runs.

Authors: We agree that the optimization details require explicit description. The pipeline minimizes a loss that directly penalizes residual timelike, spacelike, and spacetime correlation mismatches between the PAEMS-generated synthetic data and the measured repetition-code statistics; the loss is a weighted sum of these correlation residuals. Hyperparameters were selected by a modest grid search performed on a small random subset of the IBM runs, with no explicit regularization term. Because the model is intended to characterize device-specific noise for downstream synthetic-data generation, we did not enforce a formal train/validation split on the fitting data. We will revise §3 to document the exact loss function, optimizer, and hyperparameter procedure. To address generalizability concerns, we will also add a supplementary transfer experiment in which parameters fitted on one IBM device are evaluated on held-out circuit families from the same device. revision: partial
Referee: [§4] §4 (results and comparisons): The accuracy gains versus SI1000 (58–73%) and the cross-platform claims are presented without explicit held-out circuit families, ablation of the qubit-wise separation component, or transfer tests on independent datasets. The absence of these controls leaves open the possibility that the reported improvements are partly due to post-hoc fitting rather than a generalizable model.

Authors: We acknowledge the value of additional controls. The reported cross-platform results already constitute a limited transfer test, as parameters were optimized exclusively on IBM repetition-code data yet evaluated on independent QPUs (China Mobile Wuyue and China Telecom Tianyan) that were never seen during fitting. We will add (i) an ablation that isolates the contribution of the qubit-wise separation and leakage-propagation modules, (ii) performance metrics on held-out circuit families (surface-code patches and random Clifford circuits) drawn from the same IBM devices but excluded from the optimization, and (iii) a direct comparison of correlation residuals on these held-out sets. These additions will be placed in a new subsection of §4. revision: yes

Circularity Check

1 steps flagged

End-to-end optimization on repetition-code datasets used for both parameter extraction and reported correlation reductions

specific steps

fitted input called prediction [Abstract]
"Utilizing repetition-code experiment datasets, PAEMS effectively identifies the intrinsic qubit errors through an end-to-end optimization pipeline. Experiments on IBM's QPUs have demonstrated a 19.5×, 9.3×, and 5.2× reduction in timelike, spacelike, and spacetime error correlation, respectively, surpassing all of the previous works. It also outperforms the accuracy of Google's SI1000 error model by 58∼73% on multiple quantum platforms"

Parameters are extracted by optimization on the repetition-code datasets; the quoted reductions in error correlations and accuracy gains are then computed by applying those fitted parameters back to the identical datasets, making the performance numbers a direct consequence of the fit rather than an independent test.

full rationale

The paper's central performance claims rely on fitting intrinsic error parameters (including leakage) via end-to-end optimization directly to the repetition-code experiment datasets, then measuring timelike/spacelike/spacetime correlation reductions and accuracy gains against SI1000 on those same datasets. This reduces the reported 'reductions' and 'outperformance' to outcomes of the fitting process itself rather than independent predictions or held-out validation. No explicit separation between training data and evaluation circuits is described in the provided text, producing partial circularity in the evaluation chain while leaving room for cross-platform tests to provide some independent signal.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The model relies on standard quantum error propagation assumptions and introduces several fitted parameters tuned to experimental repetition-code data.

free parameters (1)

qubit-specific error rates and leakage parameters
Obtained through end-to-end optimization on repetition-code datasets; exact values not stated in abstract.

axioms (1)

standard math Errors propagate according to standard quantum mechanics on superconducting qubits
Implicit in any qubit error model; invoked when defining the separation framework.

pith-pipeline@v0.9.0 · 5525 in / 1331 out tokens · 58375 ms · 2026-05-14T00:03:34.063434+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

PAEMS adopts a physics-informed parameterization based on platform-calibrated quantities... optimized through the covariance matrix adaptation evolution strategy (CMA-ES)
IndisputableMonolith/Foundation/RealityFromDistinction.lean reality_from_one_distinction unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

Experiments on IBM's QPUs have demonstrated a 19.5×, 9.3×, and 5.2× reduction in timelike, spacelike, and spacetime error correlation

What do these tags mean?

matches: The paper's claim is directly supported by a theorem in the formal canon.
supports: The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends: The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses: The paper appears to rely on the theorem as machinery.
contradicts: The paper's claim conflicts with a theorem or certificate in the canon.
unclear: Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

50 extracted references · 50 canonical work pages · 1 internal anchor

[1]

As a special note, for single round repetition code ex- periments, correlation based metrics are not available due to the absence of temporal and spatial error accumula- tion

Individual qubit parameters may drift by a factor of 5×. As a special note, for single round repetition code ex- periments, correlation based metrics are not available due to the absence of temporal and spatial error accumula- tion. In this case, core qubit parameters are normalized and jointly optimized within a unified optimization pro- cedure. Leakage ...

work page arXiv
[2]

Calderbank, A. R. & Shor, P. W. Good quantum error- correcting codes exist.Physical Review A54, 1098 (1996)

work page 1996
[3]

G., Mariantoni, M., Martinis, J

Fowler, A. G., Mariantoni, M., Martinis, J. M. & Cleland, A. N. Surface codes: Towards practical large-scale quan- tum computation.Physical Review A—Atomic, Molecu- lar, and Optical Physics86, 032324 (2012)

work page 2012
[4]

D.et al.Coherent quantum annealing in a programmable 2,000 qubit Ising chain.Nature Physics 18, 1324–1328 (2022)

King, A. D.et al.Coherent quantum annealing in a programmable 2,000 qubit Ising chain.Nature Physics 18, 1324–1328 (2022)

work page 2022
[5]

& Zeng, W

Giurgica-Tiron, T., Hindy, Y., LaRose, R., Mari, A. & Zeng, W. J. Digital zero noise extrapolation for quantum error mitigation. In2020 IEEE International Conference on Quantum Computing and Engineering (QCE), 306– 316 (IEEE, 2020)

work page 2020
[6]

Suppressing quantum errors by scaling a surface code logical qubit.Nature614, 676– 681 (2023)

Google Quantum AI. Suppressing quantum errors by scaling a surface code logical qubit.Nature614, 676– 681 (2023)

work page 2023
[7]

Quantum error correction below the surface code threshold.Nature638, 920–926 (2025)

Google Quantum AI. Quantum error correction below the surface code threshold.Nature638, 920–926 (2025)

work page 2025
[8]

Lacroix, N.et al.Scaling and logic in the color code on a superconducting quantum processor.Nature1–3 (2025)

work page 2025
[9]

Gao, D.et al.Establishing a new benchmark in quantum computational advantage with 105-qubit zu- chongzhi 3.0 processor.Phys. Rev. Lett.134, 090601 (2025). URLhttps://link.aps.org/doi/10.1103/ PhysRevLett.134.090601

work page 2025
[10]

AbuGhanem, IBM Quantum Computers: Evolution, Performance, and Future Directions, The Journal of Supercom- puting81, 687 (2025), arXiv:2410.00916 [quant-ph]

AbuGhanem, M. Ibm quantum computers: evolu- tion, performance, and future directions.arXiv preprint arXiv:2410.00916(2024). 9

work page arXiv 2024
[11]

China un- veils record-breaking 504-qubit superconducting quan- tum computer.https://english.cas.cn/newsroom/ cas_media/202412/t20241206_893281.shtml

Chinese Academy of Sciences (CAS). China un- veils record-breaking 504-qubit superconducting quan- tum computer.https://english.cas.cn/newsroom/ cas_media/202412/t20241206_893281.shtml. Accessed Nov 2, 2025

work page 2025
[12]

Mohseni, M.et al.How to build a quantum supercom- puter: Scaling from hundreds to millions of qubits.arXiv preprint arXiv:2411.10406(2025)

work page arXiv 2025
[13]

& Preskill, J

Dennis, E., Kitaev, A., Landahl, A. & Preskill, J. Topological quantum memory.Journal of Mathematical Physics43, 4452–4505 (2002)

work page 2002
[14]

& Campbell, E

Caune, L., Reid, B., Camps, J. & Campbell, E. Be- lief propagation as a partial decoder.arXiv preprint arXiv:2306.17142(2023)

work page arXiv 2023
[15]

& Lu, C.-C

Yuan, Y. & Lu, C.-C. A modified MWPM decoding algo- rithm for quantum surface codes over depolarizing chan- nels. In2022 International Symposium on Information Theory and Its Applications (ISITA), 204–208 (IEEE, 2022)

work page 2022
[16]

& Beverland, M

Delfosse, N., Londe, V. & Beverland, M. E. To- ward a union-find decoder for quantum LDPC codes. IEEE Transactions on Information Theory68, 3187– 3199 (2022)

work page 2022
[17]

Barber, B.et al.A real-time, scalable, fast and resource- efficient decoder for a quantum computer.Nature Elec- tronics8, 84–91 (2025)

work page 2025
[18]

Bausch, J.et al.Learning high-accuracy error decoding for quantum processors.Nature635, 834–840 (2024)

work page 2024
[19]

& Vargo, A

Bravyi, S., Suchara, M. & Vargo, A. Efficient algorithms for maximum likelihood decoding in the surface code. Physical Review A90, 032326 (2014)

work page 2014
[20]

& Zhong, H.- S

Hu, G., Ouyang, W., Lu, C.-Y., Lin, C. & Zhong, H.- S. Efficient and universal neural-network decoder for stabilizer-based quantum error correction.arXiv preprint arXiv:2502.19971(2025)

work page arXiv 2025
[21]

M., Serra-Peralta, M., Byfield, D

Varbanov, B. M., Serra-Peralta, M., Byfield, D. & Terhal, B. M. Neural network decoder for near-term surface-code experiments.Physical Review Research7, 013029 (2025)

work page 2025
[22]

W., Babaie, M

Overwater, R. W., Babaie, M. & Sebastiano, F. Neural- network decoders for quantum error correction using sur- face codes: A space exploration of the hardware cost- performance tradeoffs.IEEE Transactions on Quantum Engineering3, 1–19 (2022)

work page 2022
[23]

Lange, M.et al.Data-driven decoding of quantum error correcting codes using graph neural networks.Physical Review Research7, 023181 (2025)

work page 2025
[24]

Nano Futures7, 032003 (2023)

Battistel, F.et al.Real-time decoding for fault-tolerant quantum computing: Progress, challenges and outlook. Nano Futures7, 032003 (2023)

work page 2023
[25]

Stim: a fast stabilizer circuit simulator

Gidney, C. Stim: a fast stabilizer circuit simulator. Quantum5, 497 (2021)

work page 2021
[26]

Data for ”Suppressing quantum errors by scaling a surface code logical qubit” (2022)

Google Quantum AI. Data for ”Suppressing quantum errors by scaling a surface code logical qubit” (2022). URLhttps://doi.org/10.5281/zenodo.6804040

work page doi:10.5281/zenodo.6804040 2022
[27]

Wallman, J. J. & Emerson, J. Noise tailoring for scalable quantum computation via randomized compiling.Phys- ical Review A94, 052325 (2016)

work page 2016
[28]

Salari, A.Microwave techniques in superconducting quan- tum computers(Artech House, 2024)

work page 2024
[29]

& Karpuzcu, U

Resch, S. & Karpuzcu, U. R. Benchmarking quantum computers and the impact of quantum noise.ACM Com- puting Surveys (CSUR)54, 1–35 (2021)

work page 2021
[30]

& Dutton, Z

Greenbaum, D. & Dutton, Z. Modeling coherent errors in quantum error correction.Quantum Science and Tech- nology3, 015007 (2017)

work page 2017
[31]

Feng, G.et al.Estimating the coherence of noise in quan- tum control of a solid-state qubit.Physical Review Let- ters117, 260501 (2016)

work page 2016
[32]

P., Trout, C

Barnes, J. P., Trout, C. J., Lucarelli, D. & Clader, B. Quantum error-correction failure distributions: Compar- ison of coherent and stochastic error models.Physical Review A95, 062338 (2017)

work page 2017
[33]

& Peard, N

Bravyi, S., Englbrecht, M., K¨ onig, R. & Peard, N. Cor- recting coherent errors with surface codes.npj Quantum Information4, 55 (2018)

work page 2018
[34]

E., Tarasinski, B

O’Brien, T. E., Tarasinski, B. & DiCarlo, L. Density- matrix simulation of small surface codes under current and projected experimental noise.npj Quantum Infor- mation3, 39 (2017)

work page 2017
[35]

& Gottesman, D

Aaronson, S. & Gottesman, D. Improved simulation of stabilizer circuits.Physical Review A—Atomic, Molecu- lar, and Optical Physics70, 052328 (2004)

work page 2004
[36]

The Heisenberg Representation of Quantum Computers

Gottesman, D. The Heisenberg representation of quantum computers.arXiv preprint quant-ph/9807006 (1998)

work page internal anchor Pith review Pith/arXiv arXiv 1998
[37]

& Geller, M

Katabarwa, A. & Geller, M. R. Logical error rate in the pauli twirling approximation.Scientific Reports5, 14670 (2015)

work page 2015
[38]

& Svore, K

Tomita, Y. & Svore, K. M. Low-distance surface codes under realistic quantum noise.Physical Review A90, 062320 (2014)

work page 2014
[39]

Geller, M. R. & Zhou, Z. Efficient error models for fault- tolerant architectures and the pauli twirling approxima- tion.Physical Review A—Atomic, Molecular, and Opti- cal Physics88, 012314 (2013)

work page 2013
[40]

& Broughton, M

Gidney, C., Newman, M., Fowler, A. & Broughton, M. A fault-tolerant honeycomb memory.Quantum5, 605 (2021)

work page 2021
[41]

& Almudever, C

Varsamopoulos, S., Bertels, K. & Almudever, C. G. Com- paring neural network based decoders for the surface code.IEEE Transactions on Computers69, 300–311 (2019)

work page 2019
[42]

IBM Quantum platform.https:// quantum.cloud.ibm.com/

IBM Quantum. IBM Quantum platform.https:// quantum.cloud.ibm.com/. Accessed Aug 7, 2025

work page 2025
[43]

China Mobile quan- tum computing platform.https://ecloud.10086.cn/ portal/product/WYQCLOUD/

China Mobile Research Institute. China Mobile quan- tum computing platform.https://ecloud.10086.cn/ portal/product/WYQCLOUD/. Accessed Oct 31, 2025

work page 2025
[44]

China Telecom quantum computing plat- form.https://qc.zdxlz.com/home?lang=zh/

China Telecom. China Telecom quantum computing plat- form.https://qc.zdxlz.com/home?lang=zh/. Accessed Oct 31, 2025

work page 2025
[45]

Exponential suppression of bit or phase errors with cyclic error correction.Nature595, 383–387 (2021)

Google Quantum AI. Exponential suppression of bit or phase errors with cyclic error correction.Nature595, 383–387 (2021)

work page 2021
[46]

S., Fowler, A

Wang, D. S., Fowler, A. G., Stephens, A. M. & Hollen- berg, L. C. Threshold error rates for the toric and surface codes.arXiv preprint arXiv:0905.0531(2009)

work page arXiv 2009
[47]

M., Chow, J

Gambetta, J. M., Chow, J. M. & Steffen, M. Building logical qubits in a superconducting quantum computing system.npj Quantum Information3, 2 (2017)

work page 2017
[48]

C., Wood, C

McKay, D. C., Wood, C. J., Sheldon, S., Chow, J. M. & Gambetta, J. M. Efficient z gates for quantum comput- ing.Physical Review A96, 022330 (2017)

work page 2017
[49]

M¨ uller, C., Cole, J. H. & Lisenfeld, J. Towards under- standing two-level-systems in amorphous solids: insights from quantum circuits.Reports on Progress in Physics 82, 124501 (2019)

work page 2019
[50]

V.et al.Fluctuations of energy-relaxation times in superconducting qubits.Physical Review Letters 10 121, 090502 (2018)

Klimov, P. V.et al.Fluctuations of energy-relaxation times in superconducting qubits.Physical Review Letters 10 121, 090502 (2018). Acknowledgements This work was supported by the National Natural Sci- ence Foundation of China (Grant No. 62271129) and the Sichuan Provincial Central Government Guiding Local Science and Technology Development Program (Grant...

work page 2018