pith. sign in

arxiv: 2606.29638 · v2 · pith:JCCJ5HLDnew · submitted 2026-06-28 · 🪐 quant-ph

Characterization of Unlearnable Noise with Mid-Circuit-Measurement-Based Cycle Benchmarking

M. H. Cheng , Stefano Mangini , V. Bartsch , A. C. Medina , Sergey N. Filippov , Matteo A. C. Rossi , M. S. Kim This is my paper

Pith reviewed 2026-07-01 06:19 UTC · model grok-4.3

classification 🪐 quant-ph
keywords noise characterizationmid-circuit measurementscycle benchmarkingClifford gatesPauli noisequantum error mitigationnon-Markovian noiseSPAM errors
0
0 comments X

The pith

Inserting mid-circuit measurements reverses Pauli cycles induced by Clifford gates and makes their noise components learnable.

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

The paper establishes that a generalized cycle benchmarking protocol using mid-circuit measurements and classical post-processing can identify Pauli fidelities and non-Markovian noise that standard cycle benchmarking leaves coupled and unresolvable. This matters because multi-qubit Clifford operations are central to error correction yet their full noise models have remained incomplete under SPAM errors. The key step is showing that mid-circuit measurements reverse the Pauli cycles generated by any Clifford gate, which directly yields a learnability condition. Numerical simulations and experiments on superconducting processors confirm that the added measurements disambiguate the previously inaccessible parameters when state preparation is adequate. The work also records measurement-induced bit-flip bias and non-Markovian correlations that limit the range of the Pauli noise model itself.

Core claim

An insertion of mid-circuit measurements can reverse Pauli cycles induced by a general Clifford gate. This fact enables the revelation of a Pauli-noise learnability condition for Clifford gates and renders otherwise unidentifiable Pauli fidelities and non-Markovian noise learnable via repeated measurements and classical post-processing.

What carries the argument

mid-circuit-measurement-based generalized cycle benchmarking, which applies the deferred feed-forward principle to reverse Pauli cycles and isolate previously coupled error parameters.

If this is right

  • Previously unlearnable Pauli error components in n-qubit Clifford gates become identifiable.
  • The protocol isolates non-Markovian noise in addition to Markovian Pauli fidelities.
  • Measurement-induced bit-flip bias and non-Markovian correlations define the applicability range of the Pauli noise model.
  • The method is validated on superconducting hardware and benchmarked against conventional tomography.

Where Pith is reading between the lines

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

  • The same reversal principle could be tested on non-Clifford gates to check whether the learnability condition generalizes.
  • Full noise models obtained this way would directly feed into more accurate calibration routines for error-corrected circuits.
  • Observed non-Markovian correlations suggest the protocol could be extended to quantify memory effects across multiple gate cycles.

Load-bearing premise

The protocol assumes sufficient state preparation quality to resolve the previously unlearnable noise components.

What would settle it

Running the protocol on a Clifford gate whose noise parameters remain coupled and unresolvable after mid-circuit measurements, when checked against full tomography, would show the reversal does not produce the claimed learnability.

Figures

Figures reproduced from arXiv: 2606.29638 by A. C. Medina, Matteo A. C. Rossi, M. H. Cheng, M. S. Kim, Sergey N. Filippov, Stefano Mangini, V. Bartsch.

Figure 1
Figure 1. Figure 1: Pauli noise learning of λPβ of noisy Clifford gate Λ ◦ G (cyan) with MCMs with classical feed-forwards (purple). (a) Traditional CB protocol for learning Λ suffers from unintended noise coupling induced by G that transforms n-qubit Pauli operator Pα (red) to Pβ (light pink) as Pα propagates to the measurements for final post-processing (orange). (b) By inserting single-qubit Clifford gates and partial MCMs… view at source ↗
Figure 2
Figure 2. Figure 2: CNOT/SWAP gates pattern transfer graph under CB with MCMs. A Pauli transfer graph captures the structure [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Deviation from binomial distribution in the pres [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Unlearnable noise characterization of noisy CNOT gates using mid-circuit measurements. Characterization with [PITH_FULL_IMAGE:figures/full_fig_p011_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Unlearnable noise approximation of noisy CNOT gates using MCMs with unreliable [PITH_FULL_IMAGE:figures/full_fig_p012_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Three-qubit ladder CNOT gate coefficient characterization with [PITH_FULL_IMAGE:figures/full_fig_p013_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Three CB set for decoupling unlearnable coefficients using MCM noise with layer [PITH_FULL_IMAGE:figures/full_fig_p015_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Cross quantum devices comparisons of CNOT Pauli noise learning utilizing MCMs-based CB (blue) with [PITH_FULL_IMAGE:figures/full_fig_p015_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Binomial analysis on ibm-aachen and ibm-pittsburgh that demonstrates multiple modes of non-Markovian noise contributions. (a) and (b) represent binomial analysis of ibm-aachen with/without twirling with R = 250 and S = 100. (c) and (d) represent untwirled binomial analysis of ibm-pittsburgh with/without single-qubit rotation with R = 64 and S = 3000 for the MCMs and R = 1920 and S = 100 for the CNOT gates.… view at source ↗
Figure 10
Figure 10. Figure 10: Length distribution of the MCMs read-out se [PITH_FULL_IMAGE:figures/full_fig_p018_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Fidelity difference between the approximation cal [PITH_FULL_IMAGE:figures/full_fig_p026_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: MCM-based CB interpolation for unlearnable fidelities of qubit pair [103, 204] of [PITH_FULL_IMAGE:figures/full_fig_p029_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: MCM-based CB interpolation for unlearnable fidelities of qubit pair [64, 63] of [PITH_FULL_IMAGE:figures/full_fig_p029_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: MCM-based CB interpolation for unlearnable fidelities of qubit pair [107, 108] of [PITH_FULL_IMAGE:figures/full_fig_p030_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: MCM-based CB interpolation for unlearnable fidelities of qubit pair [37, 45] of [PITH_FULL_IMAGE:figures/full_fig_p030_15.png] view at source ↗
Figure 16
Figure 16. Figure 16: Time correlations between flips ˆfi = Zi−1Zi in mid-circuit measurements, corresponding to the binomial data set in [PITH_FULL_IMAGE:figures/full_fig_p031_16.png] view at source ↗
read the original abstract

Noise characterization of multi-qubit entangling Clifford operations is a key practical bottleneck for quantum error mitigation and for the calibration, validation, and optimization of quantum error-correction protocols, especially in the presence of state preparation and measurement (SPAM) errors. Although cycle benchmarking can isolate some Pauli error components, it cannot resolve the problem of coupled error parameters, which leads to unlearnable degrees of freedom even in simple noisy gates, not to mention general $n$-qubit Clifford gates. Here we introduce mid-circuit-measurement-based generalized cycle benchmarking, a framework that makes otherwise unidentifiable Pauli fidelities and non-Markovian noise learnable via repeated measurements and classical post-processing. Applying the deferred feed-forward principle to generalized cycle benchmarking, we show that an insertion of mid-circuit measurements can reverse Pauli cycles induced by a general Clifford gate. This fact enables us to reveal a Pauli-noise learnability condition for Clifford gates. Assuming sufficient state preparation quality, we numerically demonstrate the feasibility of characterizing the previously unlearnable noise components. We implement the protocol on superconducting quantum processing units and validate its effectiveness in disambiguating the coupled noise components, benchmarked against conventional tomography. Finally, we observe consistent measurement-induced bit-flip bias and non-Markovian correlations, which define a range of applicability for the Pauli noise model and the proposed noise-characterization protocol.

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.

Referee Report

1 major / 0 minor

Summary. The paper introduces mid-circuit-measurement-based generalized cycle benchmarking, which applies the deferred feed-forward principle to insert mid-circuit measurements that reverse Pauli cycles induced by general Clifford gates. This reversal is used to derive a Pauli-noise learnability condition that renders previously unidentifiable coupled Pauli fidelities and certain non-Markovian components learnable through repeated measurements and classical post-processing. Numerical simulations demonstrate feasibility assuming sufficient state-preparation quality, and the protocol is implemented on superconducting QPUs with validation against conventional tomography; the work also reports measurement-induced bit-flip bias and non-Markovian correlations that bound the applicability of the underlying Pauli noise model.

Significance. If the reversal property is shown to hold exactly under the noise conditions present in the experiments, the protocol would address a practical bottleneck in characterizing multi-qubit Clifford noise by resolving unlearnable degrees of freedom, with direct relevance to error mitigation and quantum error correction calibration. The experimental implementation and explicit identification of non-Markovian effects provide concrete grounding for the model's range of validity.

major comments (1)
  1. [Abstract] Abstract (reversal claim) and the derivation of the learnability condition: the central assertion that mid-circuit measurements reverse Pauli cycles for arbitrary Clifford gates is load-bearing for making coupled fidelities learnable. The abstract reports observation of non-Markovian correlations and measurement-induced bias, yet the reversal is presented as following from the deferred feed-forward principle applied to a fixed Pauli channel; if the derivation does not incorporate time correlations or back-action during the cycle, the learnability condition does not extend to the reported experimental regime.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their detailed review and for identifying this key point about the scope of our theoretical claims. We respond to the major comment below.

read point-by-point responses

  1. Referee: [Abstract] Abstract (reversal claim) and the derivation of the learnability condition: the central assertion that mid-circuit measurements reverse Pauli cycles for arbitrary Clifford gates is load-bearing for making coupled fidelities learnable. The abstract reports observation of non-Markovian correlations and measurement-induced bias, yet the reversal is presented as following from the deferred feed-forward principle applied to a fixed Pauli channel; if the derivation does not incorporate time correlations or back-action during the cycle, the learnability condition does not extend to the reported experimental regime.

    Authors: We agree that the derivation of the reversal property applies the deferred feed-forward principle to a fixed Pauli channel under the standard Markovian Pauli noise model and does not incorporate time correlations or measurement back-action. This is the framework in which the learnability condition for coupled fidelities is derived. The manuscript explicitly reports the experimental observation of measurement-induced bit-flip bias and non-Markovian correlations in order to bound the regime in which the Pauli noise model (and thus the protocol) remains applicable. Numerical demonstrations assume sufficient state-preparation quality, and experimental results are validated against tomography only where the model holds. We do not assert that the learnability condition extends outside this model. To address the concern, we will revise the abstract to state more explicitly that the reversal and learnability results are obtained under the Pauli noise assumption, with experimental deviations reported separately to delineate applicability. revision: partial

Circularity Check

0 steps flagged

No circularity: learnability condition derived from deferred feed-forward principle without reduction to inputs or self-citations

full rationale

The abstract derives the reversal of Pauli cycles and the resulting Pauli-noise learnability condition by applying the deferred feed-forward principle to generalized cycle benchmarking. No equations, fitted parameters, or self-citations are shown to reduce the central claim to a tautology or prior author result by construction. The protocol is presented as an extension that makes unidentifiable components learnable via new measurements and post-processing, with applicability bounded by observed non-Markovian effects rather than assumed in the derivation itself. This matches the default expectation of a self-contained derivation against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the domain assumption of sufficient state preparation quality and the mathematical property that mid-circuit measurements reverse Pauli cycles; no free parameters or invented entities are introduced in the abstract.

axioms (1)
  • domain assumption Sufficient state preparation quality
    Required to numerically demonstrate feasibility of characterizing unlearnable components.

pith-pipeline@v0.9.1-grok · 5807 in / 1130 out tokens · 30314 ms · 2026-07-01T06:19:22.235569+00:00 · methodology

Review history (2 revisions) →

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

46 extracted references · 7 canonical work pages · 1 internal anchor

  1. [1]

    With this premise, the exact learn- ing protocol that we have been carefully developing is no longer applicable for isolating the measurement errors from the quantum gate errors

    Approximate Learning In approximate noise learning, the problem of inter- est is whether noisy MCMs may still be a reliable state preparation resource for approximating unlearnable noise pairs, when the state-projection quality of the|1⟩state is no longer guaranteed. With this premise, the exact learn- ing protocol that we have been carefully developing i...

  2. [2]

    The task is to un- ravel the unlearnable noise pair{λ 0z, λzz }, which is only learnable as a product in standard CB

    Unlearnable noise In the first experiment, the generalized CB procedure introduced in Section II is applied directly to character- ize CNOT gates, on two generations of quantum devices, ibm-aachenandibm-pittsburgh. The task is to un- ravel the unlearnable noise pair{λ 0z, λzz }, which is only learnable as a product in standard CB. To assess the quality of...

    1920

  3. [3]

    They can be learned independently from the MCMs CB set and the CNOT-MCMs set

    Classical Read-out Error In our characterization protocol, classical read-out er- rors can be learnt as the statistical difference between post-processed and non-post-processed data. They can be learned independently from the MCMs CB set and the CNOT-MCMs set. Although the detected classical read-out errors appear to be small in our experiment, they are i...

  4. [4]

    Binomial distribution In the second experiment, we apply the binomial anal- ysis to detect non-Markovian quantum effects, a key chal- lenge in quantum devices. Instead of performing bi- nomial expansion on the unlearnable pairs{λ 0z, λzz }, in this study, we perform binomial analysis directly on thelearnablefidelityλ z0, by expandingλ N z0 according to Eq...

    1920

  5. [5]

    The first type of asymmetry is the afore- mentioned zig-zag distribution that appears in panels (a) and (c)

    Asymmetries In Figure 9, we illustrate two types of asymmetry that emerge in panels (a), (c), and (d), when CB is performed on MCMs only. The first type of asymmetry is the afore- mentioned zig-zag distribution that appears in panels (a) and (c). It is characterized by its bias toward the even flips. In Figure 10, we simulate this effect by adding classic...

  6. [6]

    Based on previous studies [35], the emergence of exponen- tial tails finds a coherent explanation with the read-out dynamics when a leakage event takes place

    Non-Markovian Contributions We propose a leakage model that explains the abnor- mal exponential tail observed in the binomial analysis on the quantum gates ofibm-aachenandibm-pittsburgh. Based on previous studies [35], the emergence of exponen- tial tails finds a coherent explanation with the read-out dynamics when a leakage event takes place. With super-...

  7. [7]

    Y. Kim, A. Eddins, S. Anand, K. X. Wei, E. van den Berg, S. Rosenblatt, H. Nayfeh, Y. Wu, M. Zaletel, K. Temme, and A. Kandala, Evidence for the utility of quantum computing before fault tolerance, Nature618, 500 (2023)

    2023

  8. [8]

    Robledo-Moreno, M

    J. Robledo-Moreno, M. Motta, H. Haas, A. Javadi- Abhari, P. Jurcevic, W. Kirby, S. Martiel, K. Sharma, S. Sharma, T. Shirakawa, I. Sitdikov, R.-Y. Sun, K. J. Sung, M. Takita, M. C. Tran, S. Yunoki, and A. Mezza- capo, Chemistry beyond the scale of exact diagonaliza- tion on a quantum-centric supercomputer, Sci. Adv11, eadu9991 (2025)

    2025

  9. [10]

    van den Berg, Z

    E. van den Berg, Z. K. Minev, A. Kandala, and K. Temme, Probabilistic error cancellation with sparse pauli–lindblad models on noisy quantum processors, Nat. Phys.19, 1116–1121 (2023)

    2023

  10. [11]

    Filippov, M

    S. Filippov, M. Leahy, M. A. C. Rossi, and G. Garc´ ıa- P´ erez, Scalable tensor-network error mitigation for near- term quantum computing (2023), arXiv:2307.11740

    work page arXiv 2023

  11. [12]

    S. N. Filippov, S. Maniscalco, and G. Garc´ ıa-P´ erez, Scal- ability of quantum error mitigation techniques: From utility to advantage (2024), arXiv:2403.13542

    work page arXiv 2024

  12. [13]

    Mangini, M

    S. Mangini, M. Cattaneo, D. Cavalcanti, S. Filippov, M. A. C. Rossi, and G. Garc´ ıa-P´ erez, Tensor network noise characterization for near-term quantum computers, Phys. Rev. Res.6, 033217 (2024)

    2024

  13. [14]

    L. E. Fischer, M. Leahy, A. Eddins, N. Keenan, D. Fer- racin, M. A. C. Rossi, Y. Kim, A. He, F. Pietracaprina, B. Sokolov, S. Dooley, Z. Zimbor´ as, F. Tacchino, S. Man- iscalco, J. Goold, G. Garc´ ıa-P´ erez, I. Tavernelli, A. Kan- dala, and S. N. Filippov, Dynamical simulations of many- body quantum chaos on a quantum computer, Nat. Phys. 22, 302 (2026)

    2026

  14. [15]

    C. H. Bennett, D. P. DiVincenzo, J. A. Smolin, and W. K. Wootters, Mixed-state entanglement and quantum error correction, Phys. Rev. A54, 3824–3851 (1996)

    1996

  15. [16]

    Onorati, T

    E. Onorati, T. Kohler, and T. S. Cubitt, Fitting quan- tum noise models to tomography data, Quantum7, 1197 (2023)

    2023

  16. [17]

    Jayakumar, S

    A. Jayakumar, S. Chessa, C. Coffrin, A. Y. Lokhov, M. Vuffray, and S. Misra, Universal framework for simul- taneous tomography of quantum states and spam noise, Quantum8, 1426 (2024)

    2024

  17. [18]

    Erhard, J

    A. Erhard, J. J. Wallman, L. Postler, M. Meth, R. Stricker, E. A. Martinez, P. Schindler, T. Monz, J. Emerson, and R. Blatt, Characterizing large-scale quantum computers via cycle benchmarking, Nat. Com- mun.10, 5347 (2019)

    2019

  18. [19]

    Nielsen, J

    E. Nielsen, J. K. Gamble, K. Rudinger, T. Scholten, K. Young, and R. Blume-Kohout, Gate set tomography, Quantum5, 557 (2021)

    2021

  19. [20]

    S. Chen, Y. Liu, M. Otten, A. Seif, B. Fefferman, and L. Jiang, The learnability of Pauli noise, Nat. Commun. 14, 269 (2023)

    2023

  20. [21]

    Bluvstein, S

    D. Bluvstein, S. J. Evered, A. A. Geim, S. H. Li, H. Zhou, T. Manovitz, S. Ebadi, M. Cain, M. Kali- nowski, D. Hangleiter, J. P. Bonilla Ataides, N. Maskara, I. Cong, X. Gao, P. Sales Rodriguez, T. Karolyshyn, G. Semeghini, M. J. Gullans, M. Greiner, V. Vuleti´ c, and M. D. Lukin, Logical quantum processor based on reconfigurable atom arrays, Nature626, 58 (2024)

    2024

  21. [22]

    Y. Yu, K. Yan, D. Biswas, V. N. Zhang, B. Harraz, C. Noel, C. Monroe, and A. Kozhanov, In situ mid- circuit qubit measurement and reset in a single-species trapped-ion quantum computing system, Phys. Rev. Res. 7, 043355 (2025)

    2025

  22. [23]

    Bengtsson, A

    A. Bengtsson, A. Opremcak, M. Khezri, D. Sank, A. Bourassa, K. J. Satzinger, S. Hong, C. Erickson, B. J. Lester, K. C. Miao, A. N. Korotkov, J. Kelly, Z. Chen, and P. V. Klimov, Model-based optimization of super- conducting qubit readout, Phys. Rev. Lett.132, 100603 (2024)

    2024

  23. [24]

    V. A. Zhuravlev and S. N. Filippov, Quantum state to- mography via sequential uses of the same informationally incomplete measuring apparatus, Lobachevskii Journal of Mathematics41, 2405 (2020)

    2020

  24. [25]

    I. A. Luchnikov and S. N. Filippov, Quantum evolution in the stroboscopic limit of repeated measurements, Phys. Rev. A95, 022113 (2017)

    2017

  25. [26]

    Grimaudo, A

    R. Grimaudo, A. Messina, A. Sergi, N. V. Vitanov, and S. N. Filippov, Two-qubit entanglement genera- tion through non-Hermitian Hamiltonians induced by re- peated measurements on an ancilla, Entropy22, 1184 (2020)

    2020

  26. [27]

    A. D. C´ orcoles, M. Takita, K. Inoue, S. Lekuch, Z. K. Minev, J. M. Chow, and J. M. Gambetta, Exploiting dynamic quantum circuits in a quantum algorithm with superconducting qubits, Phys. Rev. Lett.127, 100501 (2021)

    2021

  27. [28]

    Lemelin, C

    A. Lemelin, C. Pere, O. Landon-Cardinal, and C. Coti, Mid-circuit measurement as an algorithmic primitive (2025), arXiv:2506.00118

    work page arXiv 2025

  28. [29]

    Zhang, S

    Z. Zhang, S. Chen, Y. Liu, and L. Jiang, Generalized cy- cle benchmarking algorithm for characterizing midcircuit measurements, PRX Quantum6, 010310 (2025). 21

    2025

  29. [30]

    Hines and T

    J. Hines and T. Proctor, Pauli noise learning for mid- circuit measurements, Phys. Rev. Lett.134, 020602 (2025)

    2025

  30. [31]

    Hothem, J

    D. Hothem, J. Hines, C. Baldwin, D. Gresh, R. Blume- Kohout, and T. Proctor, Measuring error rates of mid- circuit measurements, Nat. Commun.16, 5761 (2025)

    2025

  31. [32]

    Introduction to Quantum Gate Set Tomography

    D. Greenbaum, Introduction to quantum gate set tomog- raphy (2015), arXiv:1509.02921

    work page internal anchor Pith review Pith/arXiv arXiv 2015

  32. [33]

    Z. Cai, R. Babbush, S. C. Benjamin, S. Endo, W. J. Hug- gins, Y. Li, J. R. McClean, and T. E. O’Brien, Quantum error mitigation, Rev. Mod. Phys.95, 045005 (2023)

    2023

  33. [34]

    M. R. Geller and Z. Zhou, Efficient error models for fault- tolerant architectures and the pauli twirling approxima- tion, Phys. Rev. A88, 012314 (2013)

    2013

  34. [35]

    S. T. Flammia and J. J. Wallman, Efficient estimation of Pauli channels, ACM Trans. Quantum Comput.1, 3 (2020)

    2020

  35. [36]

    S. J. Beale and J. J. Wallman, Randomized compiling for subsystem measurements (2023), arXiv:2304.06599

    work page arXiv 2023

  36. [37]

    McLaren, M

    D. McLaren, M. A. Graydon, and J. J. Wall- man, Stochastic errors in quantum instruments (2023), arXiv:2306.07418

    work page arXiv 2023

  37. [38]

    A. H. Myerson, D. J. Szwer, S. C. Webster, D. T. C. Allcock, M. J. Curtis, G. Imreh, J. A. Sherman, D. N. Stacey, A. M. Steane, and D. M. Lucas, High-fidelity readout of trapped-ion qubits, Phys. Rev. Lett.100, 200502 (2008)

    2008

  38. [39]

    Y. Quek, D. Stilck Fran¸ ca, S. Khatri, J. J. Meyer, and J. Eisert, Exponentially tighter bounds on limitations of quantum error mitigation, Nat. Phys.20, 1648 (2024), pMID: 39416853; PMCID: PMC11473368

    2024

  39. [40]

    E. H. Chen, S. Chen, L. E. Fischer, A. Eddins, L. C. G. Govia, B. Mitchell, A. He, Y. Kim, L. Jiang, and A. Seif, Disambiguating pauli noise in quantum comput- ers (2025), arXiv:2505.22629

    work page arXiv 2025

  40. [41]

    Hazra, W

    S. Hazra, W. Dai, T. Connolly, P. D. Kurilovich, Z. Wang, L. Frunzio, and M. H. Devoret, Benchmark- ing the readout of a superconducting qubit for repeated measurements, Phys. Rev. Lett.134, 100601 (2025). Appendix A: Mathematical Foundations of MCM Noise Learning In this section, we establish key results in MCM noise learning for generaln-qubit Clifford ...

    2025

  41. [42]

    Pauli Propagation of Stochastic CNOT Theorem 5(Pauli Propagation for Stochastic CNOT). A stochastic CNOT, formed by tracing out the classical register of the QICF, replicates the Pauli transformation of a coherent CNOT gate onZ-type observables, up to a scalar prefactor from the classical readout error. For- mally, tracing out the classical register gives...

  42. [43]

    Learnable Degrees of Freedom Theorem 7.For cycles of quantum instruments, the classical read-out error is learnable SPAM-free if the pre- measurement and post-measurement errors are separable. Proof.Based on Theorems 2 and 5, the following can be learned via CB, isolating SPAM errors (pandqare dropped given the clear context): f1 =A(ω ZcZωZc+1Z)n +B,(A19)...

  43. [44]

    Lemma 1(Deferred Feed-Forward Principle for CB)

    Deferred Feed-Forward Principle for Generalized CB Building on the learnability results for Markovian MCM errors, we establish the second key result in this work — that single-qubit Clifford gates, combined with MCMs, are sufficient to decouple all unlearnable degrees of freedom for arbitraryn-qubit Clifford gates. Lemma 1(Deferred Feed-Forward Principle ...

  44. [45]

    In this definition, the eventsa andbare separable

    Binomial Distribution Consider a more restricted form of USI — a separable l= 1 USI, acting on qubitp: Nk(·) = X a,b∈Z2 pap′ b ΛaΛ′ b ⊗ X b p |k⟩⟩⟨⟨k|pX a p K (·).(C3) We adopt this representation to isolate a single-qubit measurement event from a parallel operation of MCMs governed by Theorem 6. In this definition, the eventsa andbare separable. The nota...

  45. [46]

    X keven N0 k ϵk 0(1−ϵ 0)N0−k # ×

    Classical Read-out Error In Eq. (23), it is already assumed that the read-out er- rors are averaged due to randomized compilation. How- ever, we expect that the classical read-out assignment rate differs depending on the collapsed quantum basis state 0 or 1. In this section, we derive the effective read- out error fidelity due to randomized compilation. L...

  46. [47]

    Here, the effect of classical read-out errors on the binomial distribu- tion is further explored

    Impact on Binomial Analysis In the previous sections, we derived the binomial ex- pression for binomial analysis directly from propagating the Paulis through the channels, and computed the im- pact of classical read-out errors on the fidelity. Here, the effect of classical read-out errors on the binomial distribu- tion is further explored. From Section IV...