Pith. sign in

REVIEW 3 major objections 4 minor 2 cited by

State Similarity in Modular Superconducting Quantum Processors with Classical Communications

T0 review · 3 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read This paper claims that wire cutting turns cross-platform state verification into a task whose sample cost shrinks exponentially with the number of circuit parts.

desk verdict The r=2 core is solid and the experiments are real, but the advertised sub-exponential gain for r≥3 hangs on a one-line proof that needs to be derived. read the letter →

arxiv 2506.01657 v3 pith:TSTHQ2XU submitted 2025-06-02 quant-ph

Bujiao Wu , Changrong Xie , Peng Mi , Zhiyi Wu , Zechen Guo , Peisheng Huang , Wenhui Huang , Xuandong Sun
show 15 more authors
Jiawei Zhang Libo Zhang Jiawei Qiu Xiayu Linpeng Ziyu Tao Ji Chu Ji Jiang Song Liu Jingjing Niu Yuxuan Zhou Yuxuan Du Wenhui Ren Youpeng Zhong Tongliang Liu Dapeng Yu
This is my paper · ORCID
classification quant-ph
keywords cross-platformverificationcircuitcuttingwirefidelityestimationsamplecomplexityfederatedquantumkernelsuperconductingprocessorrandomizedmeasurements
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper asks whether the outputs of modular quantum processors connected only by classical communication can be certified, and answers yes with a sample-complexity claim. It proposes a cross-platform fidelity estimator for states prepared by wire cutting, and claims that estimating $\mathrm{Tr}(\rho\sigma)$ needs $O((2^{k_1+1}+1)^{2k_2} 6^{n/(2r)}/\varepsilon)$ measurements, where $n$ is the total qubit count, $r$ is the number of independent circuit parts, $k_1$ is the number of qubits in each cut, and $k_2$ is the number of cuts. If right, verification becomes much cheaper as a circuit is split into more modules, and for $r\ge 3$ it improves on every prior cross-platform scheme without requiring global random unitaries. The authors demonstrate the protocol on a superconducting processor, certifying 11-qubit GHZ states using modules of at most 6 qubits, and use the same machinery to build a privacy-preserving federated quantum kernel that learns a phase transition with only eight training samples.

What carries the argument

The load-bearing mechanism is the combination of wire cutting with cross-platform randomized-measurement fidelity estimation. Wire cutting replaces each cut wire by a measure-and-prepare channel sampled through a Bernoulli variable and a Clifford unitary, with weight $2^{k_1+1}+1$ per cut, splitting state preparation into $r$ independent parts. On each part, both platforms apply the same random local unitary from a tensor product of local 4-designs and measure in the computational basis; the classical post-processor forms $\hat v$ by multiplying per-part Hamming-distance kernels $(-2)^{-D(s,s')}$ and summing over cutting variables and outcomes with their weights. The key mathematical identities are the twirling property of Clifford groups and local 4-designs, which turn averaged measurement frequencies into state overlaps, and a variance bound that decouples across the $r$ parts, producing the $6^{n/(2r)}$ factor. A parallel-execution variant enumerates all cutting configurations so the subcircuits can run concurrently instead of sequentially.

What would settle it

One concrete check: build an $r=3$ circuit where two cuts feed into the same subcircuit, evaluate estimator (S41) on a fixed $n$-qubit GHZ state, and plot the number of samples needed for $|\hat v-\mathrm{Tr}(\rho\sigma)|\le\varepsilon$. If the required samples scale like $6^{n/2}/\varepsilon$ instead of the claimed $(2^{k_1+1}+1)^{2k_2}6^{n/(2r)}/\varepsilon$, the decoupling assumption behind Theorem 5 is false.

Watch

Extended reading notes

Core claim

The central claim is Theorem 1 in the main text, stated formally as Theorem 5 in the supplementary information: for an $n$-qubit state prepared by wire cutting into $r$ independent subcircuits with $k_2$ cuts each involving at most $k_1$ qubits, the proposed estimator $\hat v$ is unbiased for $\mathrm{Tr}(\rho\sigma)$, and with $O((2^{k_1+1}+1)^{2k_2} 6^{n/(2r)}/\varepsilon)$ measurements it satisfies $|\hat v-\mathrm{Tr}(\rho\sigma)|\le \varepsilon$ with high probability. The estimator applies identical local random unitaries on both platforms, weights measurement outcomes by $(-2)^{-D(s,s')}$, and sums over all cutting configurations with their sampling weights. For constant $k_1,k_2$, the paper claims this gives roughly $O(1.565^n/\varepsilon)$ samples when $r=2$, compared with the roughly $O(2^n)$ of the original distance-based protocol, and for $r\ge 3$ a sub-exponential improvement over all existing cross-platform algorithms while avoiding global random unitaries. Experimentally, it reports cross-platform fidelities for GHZ states up to 11 qubits using at most 6-qubit modules, and a federated quantum kernel that predicts the phase of a 5-qubit transverse-field Ising model with eight training examples.

Load-bearing premise

The general multi-part theorem is stated with the proof deferred as similar to Theorem 4, and it assumes the error bound derived for a chain of parts carries over unchanged when the circuit is cut into $r$ arbitrarily arranged independent parts; if overlapping cuts break that decoupling, the claimed saving for $r\ge 3$ does not follow.

Editorial extensions

If this is right

  • For two modules ($r=2$) with constant cuts, the sample count drops from the roughly $2^n$ of the prior distance-based protocol to $O(1.565^n/\varepsilon)$, and only local random unitaries are needed.
  • For $r\ge 3$ with constant $k_1,k_2$, the bound $O((2^{k_1+1}+1)^{2k_2}6^{n/(2r)}/\varepsilon)$ is a sub-exponential improvement over every prior cross-platform verification algorithm for arbitrary states.
  • The same measurement records can be reprocessed into stabilizer expectation values, so fidelity to an ideal GHZ state can be reported from the same shots, and a calibration protocol suppresses cut-qubit measurement noise.
  • Because each kernel entry is a state overlap estimated without sharing raw states, the protocol yields a federated quantum kernel; the reported 5-qubit phase-learning experiment extracts the phase boundary with only eight training examples.
  • The paper's auxiliary analysis proves that even without circuit cutting, local randomized-measurement cross-platform verification needs $\sqrt{(18/5)^n}$ copies when $\rho=\sigma$ and $\sqrt{6^n}$ when $\rho\ne\sigma$, separating it sharply from single-copy tomography's $\Omega(2^{2n})$ copy cost.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the decoupling behind Theorem 5 holds in full generality, modularity becomes a verification resource: each additional cut divides the exponent $n/(2r)$, so the same fidelity target becomes cheaper as the processor is split into more modules.
  • The federated kernel construction is a general template: any kernel method whose entries are quantum state overlaps can be run across data holders that exchange only classical measurement statistics, and testing it on multi-class classification or disordered Hamiltonians would be a natural next step.
  • Because the protocol needs only local 4-designs and classical communication, it is portable to hardware without quantum links; a direction the paper leaves open is allowing limited quantum communication between modules, which might further lower the cost.
  • A direct numerical test of the $r\ge 3$ claim would compute the variance of estimator (S41) on circuits where cuts are not arranged in a chain; the paper's simulations cover GHZ states with up to three sequential cuts, not arbitrary topologies.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper develops a cross-platform fidelity estimation protocol for modular quantum processors using wire cutting. Given two n-qubit states prepared on two platforms, the protocol applies random unitaries to each of r subcircuits after cutting and combines measurement frequencies with cutting weights to estimate Tr(rho sigma). The main theoretical claim is a sample-complexity bound O((2^{k1+1}+1)^{2k2} 6^{n/(2r)}/epsilon) for the estimator, which would give a sub-exponential improvement for r >= 3 over protocols using global random unitaries. The SI provides a full proof for the sequential 'staircase' cutting case (Theorem 4) and sketches the extension to general r-part topologies (Theorem 5). The paper also proves sample-complexity bounds for the distance-based cross-platform algorithm (SI B), reports an experiment on a superconducting processor verifying GHZ states up to 11 qubits using modules of at most 6 qubits, and demonstrates a federated quantum kernel for phase learning.

Significance. If the general r-part bound (Theorem 5) is correct, the result is significant: it establishes that cross-platform state similarity can be certified in the circuit-cutting regime with sample complexity exponential in n/r rather than n, and it provides a practical verification method for modular processors. The detailed variance analysis of the distance-based algorithm in SI B fills a theoretical gap left by numerical studies. The experimental demonstration---GHZ states up to 11 qubits from 6-qubit modules and a privacy-preserving quantum kernel---is valuable and appears carefully executed. However, the advertised r >= 3 improvement rests on an unproven variance-decoupling step, and the experimental ensemble does not meet the 4-design condition required by the theorem; these gaps materially affect the strength of the claims.

major comments (3)
  1. [SI C.3, Theorem 5 and Eq. (S40)-(S41)] The proof of Theorem 5 is not supplied. The sentence 'The proof is similar to Theorem 4 by generalizing the sequential cutting...' does not provide the required variance-decoupling argument. Theorem 4's variance bound in Eq. (S27) uses the sequential staircase structure in which each of the k2+1 parts is connected to at most two cuts and the full-outcome probability factorizes over parts. In the general r-part setting, the conditioning set c_{pre(j),z} in Eq. (S40) couples parts through shared or incoming cuts, and the sum over c,c' couples all parts. The claim that 'cross-concatenation' yields an effective m^{k2+1} (or m^r) shot count is not derived for this topology. Since the sub-exponential improvement for r >= 3 is the central advertised separation, this missing derivation is load-bearing. The authors should provide a self-contained proof of Theorem 5 or restrict the bound to the sequential case.
  2. [Main text Theorem 1 / SI C.2, Theorem 4] The sample-complexity guarantee requires the random unitaries to form a local 4-design. The experimental implementation described in SI E and Fig. 2 uses single-qubit local Clifford unitaries, which do not form a local 4-design (the qubit Clifford group is at best a 3-design). Consequently, the rigorous bound of Theorem 4 does not apply to the experiments as executed. Either the experiments should be repeated with a 4-design ensemble, or the paper should clearly state that the experimental results are heuristic demonstrations not covered by the theorem's guarantee.
  3. [Main text Eq. (2) and Fig. 2] The paper defines cross-platform fidelity F(rho,sigma) with purity normalization in Eq. (2), but the estimator in Eq. (3) and Theorem 1 estimates Tr(rho sigma), not F. The experimental values reported as 'cross-platform fidelities' in Fig. 2 appear to be unnormalized Tr(rho sigma) estimates. If so, the label is misleading for noisy states, where purities are less than one. Please either report Tr(rho sigma) as the target quantity throughout, or estimate the purities and report the normalized fidelity.
minor comments (4)
  1. [SI C.2, Eq. (S26)] The variance identity is written as Var_{W,X}(v_{z,z'}) = Var_{W,X}(v_{z,z'})/N^{k2+1}; the left-hand side and right-hand side use the same symbol for different quantities. Rename the single-combination estimator (for example, \tilde v) or write the variance of the averaged estimator explicitly.
  2. [Main text Theorem 1 and SI E] The theorem states that unbiasedness requires a local 2-design and concentration requires a local 4-design; the experimental section should state explicitly which ensemble is used and whether it satisfies the 4-design condition.
  3. [SI C.3, after Eq. (S40)] The text states 'provided that k1 and k2 are constants with k1 < r and r >= 3', but this condition is not included in the statement of Theorem 5. Please reconcile the theorem statement with the surrounding conditions.
  4. [Main text Section II.A] The phrase 'sub-exponential improvement' could be clarified: for fixed r, 6^{n/(2r)} is still exponential in n with a reduced base. Consider using 'exponential with reduced exponent' or defining 'sub-exponential' precisely to avoid ambiguity.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the estimator and its sample-complexity bound are derived from standard twirling and variance bounds, with the main weakness being an omitted proof for Theorem 5, not a circular reduction.

full rationale

The paper's central claim is that the estimator in Eq. (3) approximates Tr(ρσ) with the stated sample complexity. This is supported by an explicit derivation: SI C proves unbiasedness and gives variance bounds for the sequential cutting case using representation-theoretic lemmas (Lemma 2, Proposition 1, Lemma 3) and standard twirling identities. The target quantity Tr(ρσ) is not an input to the estimator; rather, the estimator is constructed so that its expectation equals it, which is a proof obligation, not circularity. Theorem 4 contains a concrete variance calculation, and Theorem 5 is asserted by the sentence 'The proof is similar to Theorem 4 by generalizing the sequential cutting with generating k2+1 parts into more generating cutting with generating r independent parts.' That is a missing-proof or generality gap, not a circular reduction: no equation in the paper identifies the claimed bound with an input assumption. The experimental fidelities are reported measured values, not fitted parameters renamed as predictions. Self-citations such as Refs. 46 and 80 appear in background or as measurement/application tools and are not the load-bearing justification for the main scaling claim. Therefore, no circular step is exhibited, and the derivation is self-contained up to the noted omitted proof of the general r-part variance bound.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The central derivation has no fitted parameters; the only adjustable quantities are experimental resource counts (N shots per random unitary, m repetitions, L Bernoulli samples) that set the statistical error, not the target value. The proof relies on standard representation-theoretic facts and on the domain assumption that the same local 4-design unitaries can be applied identically on both platforms. The experimental error mitigation introduces a measurement-noise model that is assumed dominant; violations would bias the reported fidelities. No new physical entities are postulated.

assumptions (5)
  • standard math The operators {M_i} and {κ_i} in Eq. (1) form an informationally complete measurement and corresponding reinitialized states for the identity channel decomposition.
    This is the wire-cutting basis from Lowe et al. (Ref. 23), cited and used throughout SI A.2.
  • domain assumption The random unitaries in the second randomness source can be sampled from a unitary local 4-design, and the same sequence is used on both platforms.
    Theorem 1 requires a local 4-design for the variance bound; the experimental implementation uses single-qubit Clifford unitaries (SI C.1), which are not a 4-design, so this assumption is not met in the experiment.
  • standard math The variance of the local distance-based estimator for arbitrary ρ,σ is bounded by 2^n + 3^n/m^2 (Proposition 2), which relies on Lemma 3 and the enumeration table in SI B.
    This bound is proven in SI B using representation theory of the symmetric group; we did not find an error, but the table is central.
  • domain assumption Measurement noise is the dominant error in the cut-qubit channel and can be modeled as a tensor product of independent two-level channels (SI C.4, Eq. S47).
    The error mitigation protocol in SI C.4 assumes state preparation and single-qubit gates on the cut qubits are noise-free; if readout errors are not dominant, the mitigation is biased.
  • domain assumption The processor regions used as 'Platform 1' and 'Platform 2' behave as independent modular processors despite being on the same chip.
    The experiment emulates modular architecture by partitioning one chip; cross-talk or shared control could break independence, but no evidence of this is reported.

how reviews work

0 comments
Cite this review

Pith. "Pith review of State Similarity in Modular Superconducting Quantum Processors with Classical Communications." pith.science (2026). https://pith.science/paper/TSTHQ2XU

@misc{pith2026250601657,
  author       = {Pith},
  title        = {Pith review of: State Similarity in Modular Superconducting Quantum Processors with Classical Communications},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TSTHQ2XU}},
  note         = {Machine review of arXiv:2506.01657}
}
read the original abstract

As quantum devices continue to scale, distributed quantum computing emerges as a promising strategy for executing large-scale tasks across modular quantum processors. A central challenge in this paradigm is verifying the correctness of computational outcomes when subcircuits are executed independently following circuit cutting. Here we propose a cross-platform fidelity estimation algorithm tailored for modular architectures. Our method achieves substantial reductions in sample complexity compared to previous approaches designed for single-processor systems. We experimentally implement the protocol on modular superconducting quantum processors with up to 6 qubits to verify the similarity of two 11-qubit GHZ states. Beyond verification, we show that our algorithm enables a federated quantum kernel method that preserves data privacy. As a proof of concept, we apply it to a 5-qubit quantum phase learning task using six 3-qubit modules, successfully extracting phase information with just eight training samples. These results establish a practical path for scalable verification and trustworthy quantum machine learning of modular quantum processors.

Figures

Figures reproduced from arXiv: 2506.01657 by the authors.

Figure 1
Figure 1. FIG. 1. An overview of state similarity estimation between two 5-qubit states using circuit cutting across 4-qubit platforms [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. Illustration for state similarity in modular superconducting quantum processors with classical communications. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3. Experimental phase learning using the cross-platform verification in circuit-cutting case. [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Optimal Distributed Similarity Estimation of Quantum Channels

    quant-ph 2025-12 reject novelty 5.0 of 10

    Settling DSEC at Θ(max{√d/ε,1/ε²}) is claimed, but the general-channel protocol's unbiasedness proof uses a false identity, and explicit non-unital channels break the estimator.

  2. Demonstration of Efficient Predictive Surrogates for Large-scale Quantum Processors

    quant-ph 2025-07 conditional novelty 5.0 of 10

    Classical surrogates using truncated trigonometric expansions emulate noisy quantum processors and cut measurement overhead in VQE pre-training and Floquet phase identification.

Reference graph

Works this paper leans on

50 extracted references · 44 canonical work pages · cited by 2 Pith papers

  1. [1]

    Distance-based algorithm for cross-platform fidelity estimation 12 b

    Cross-platform fidelity estimation 12 a. Distance-based algorithm for cross-platform fidelity estimation 12 b. Collision-based algorithm for cross-platform fidelity estimation 13

    work page

  2. [2]

    Preparation of the quantum state with circuit cutting 13

    work page

  3. [3]

    Complexity analysis of Distance-based Cross-platform fidelity estimation algorithm 15 C

    Overview of quantum kernel 14 B. Complexity analysis of Distance-based Cross-platform fidelity estimation algorithm 15 C. Algorithmic details 19

    work page

  4. [4]

    Algorithm details 27 D

    Cross-platform fidelity for noiseless and noisy graph states with measurement error mitigation 27 a. Algorithm details 27 D. Details of superconducting quantum processors 30 E. More Experimental Results 30

    work page

  5. [36]

    Implementation of wire cutting in parallel 19

    work page

  6. [37]

    Cross-platform fidelity estimation with regular cutting 22

    work page

  7. [38]

    Cross-platform fidelity estimation with more general cutting 26

    work page

  8. [40]

    Parameter settings for the GHZ-state experiments in the main text 31 b

    Experimental implementation of cross-platform fidelity estimation of GHZ states with circuit cutting 31 a. Parameter settings for the GHZ-state experiments in the main text 31 b. Experimental circuit settings in the main text 31 c. More experimental results for GHZ-states 32 d. Simulation results in multiple cuttings 33

    work page

Show all 50 references
  1. [41]

    Problem setup 35 b

    Quantum phase learning 35 a. Problem setup 35 b. Parameter settings and simulation results 35 c. Experimental circuit settings in the main file 36 12 In this Supplementary Information (SI), we provide detailed descriptions of the algorithms, theoretical analyses, experimental ...

  2. [42]

    distance-based algorithm

    Cross-platform fidelity estimation Letρandσbe two quantum states prepared by two quantum platforms, denoted by Platform 1 and Platform 2, respectively. The cross-platform fidelity [32, 55–57] between the states prepared by Platform 1 and Platform 2 is F(ρ, σ) = Tr (ρσ)p Tr (ρ2...

  3. [43]

    𝑈!𝑘𝑈" 𝑘 𝒞𝒞#𝒄 𝑈!𝑘 𝒄 𝑈

    Preparation of the quantum state with circuit cutting Lowe et al. [23] proposed the wire-cutting algorithm by introducing the random measurement technique. The core idea is to introduce two channels Φ 0(·) and Φ 1(·) operating onkqubits, such that Φ0 =E C C† ◦M s ◦ C ,(S6) whe...

  4. [44]

    Among these techniques, kernel methods play a crucial role by enabling the application of linear algorithms to nonlinear problems through the use of feature maps

    Overview of quantum kernel Machine learning techniques have become pivotal in the era of big data, demonstrating significant potential in both classical computing [59, 60] and quantum computing [34, 61–63]. Among these techniques, kernel methods play a crucial role by enabling...

  5. [45]

    To address this bottleneck, we 20 develop an improved version of the proposed algorithm

    Implementation of wire cutting in parallel Recall that the general procedure of wire cutting proceeds sequentially [23], which results in high runtime costs in practice, particularly in the proposed cross-platform fidelity estimation algorithm. To address this bottleneck, we 2...

  6. [46]

    𝑈& 𝒔&𝒄& 𝑉& 𝒔&#𝒄&# 𝑈' 𝒔'𝒄' 𝑉' 𝒔'#𝒄'# ∑𝑝'𝑞' 𝒔#,𝒔#

    Cross-platform fidelity estimation with regular cutting The state preparation circuit in the extreme case takes the form of a staircase composed of sequential sub-blocks, as illustrated in Supplementary Figure C.6a. We considerk 2 rounds of cutting, resulting ink 2 + 1 parts a...

  7. [47]

    We considerk 2 rounds of cutting, resulting inrparts after all the cuts

    Cross-platform fidelity estimation with more general cutting Now we extend Theorem 4 into a broader cutting framework, where the employed circuits are permitted to have more intricate configurations, as illustrated in Supplementary Figure C.7. We considerk 2 rounds of cutting,...

  8. [48]

    𝑈! 𝑈" 𝒞𝒞#𝑐𝑐 𝑈!𝑐 𝑈

    Cross-platform fidelity for noiseless and noisy graph states with measurement error mitigation Here we give the analysis for the fidelities between the generated noisy graph state with circuit cutting and the noiseless graph state Tr (ˆρρG), whereGdenotes the graph representat...

  9. [49]

    #W” column), (2) the number of circuits induced by the cutting process (in the “# Circ. aW

    Experimental implementation of cross-platform fidelity estimation of GHZ states with circuit cutting a. Parameter settings for the GHZ-state experiments in the main text Supplementary Table E.2 summarizes the number of random unitaries applied for each qubit configuration, alo...

  10. [50]

    derivative reduction by adiabatic gate technique

    Quantum phase learning a. Problem setup Quantum phase recognition plays a central role in understanding complex quantum systems, particularly in con- densed matter physics and quantum chemistry. In this work, we focus on learning the phase structure of the transverse- field Is...

  11. [52]

    Qubit-reuse compilation with mid-circuit mea- surement and reset

    Matthew DeCross, Eli Chertkov, Megan Kohagen, and Michael Foss-Feig. Qubit-reuse compilation with mid-circuit mea- surement and reset. Physical Review X, 13(4):041057, 2023

    2023

  12. [53]

    Enhancing dispersive readout of superconducting qubits through dynamic control of the dispersive shift: Experiment and theory

    Fran¸ cois Swiadek, Ross Shillito, Paul Magnard, Ants Remm, Christoph Hellings, Nathan Lacroix, Quentin Ficheux, Dante Colao Zanuz, Graham J Norris, Alexandre Blais, et al. Enhancing dispersive readout of superconducting qubits through dynamic control of the dispersive shift: ...

    2024

  13. [54]

    Tensor networks and graphical calculus for open quantum systems

    Christopher J Wood, Jacob D Biamonte, and David G Cory. Tensor networks and graphical calculus for open quantum systems. arXiv preprint arXiv:1111.6950, 2011

    2011 arXiv

  14. [55]

    An alternative quantum fidelity for mixed states of qudits

    Xiaoguang Wang, Chang-Shui Yu, and XX Yi. An alternative quantum fidelity for mixed states of qudits. Physics Letters A, 373(1):58–60, 2008

    2008

  15. [56]

    Quantum fidelity measures for mixed states

    Yeong-Cherng Liang, Yu-Hao Yeh, Paulo EMF Mendon¸ ca, Run Yan Teh, Margaret D Reid, and Peter D Drummond. Quantum fidelity measures for mixed states. Reports on Progress in Physics, 82(7):076001, 2019

    2019

  16. [57]

    Probing r´ enyi entanglement entropy via randomized measurements

    Tiff Brydges, Andreas Elben, Petar Jurcevic, Beno ˆ ıt Vermersch, Christine Maier, Ben P Lanyon, Peter Zoller, Rainer Blatt, and Christian F Roos. Probing r´ enyi entanglement entropy via randomized measurements. Science, 364(6437):260–263, 2019

    2019

  17. [58]

    Representation theory: a first course, volume 129

    William Fulton and Joe Harris. Representation theory: a first course, volume 129. Springer Science & Business Media, 2013

    2013

  18. [59]

    Deep learning

    Yann LeCun, Yoshua Bengio, and Geoffrey Hinton. Deep learning. Nature, 521(7553):436–444, 2015

    2015

  19. [60]

    Why does deep and cheap learning work so well? Journal of Statistical Physics, 168:1223–1247, 2017

    Henry W Lin, Max Tegmark, and David Rolnick. Why does deep and cheap learning work so well? Journal of Statistical Physics, 168:1223–1247, 2017

    2017

  20. [61]

    Solving the quantum many-body problem with artificial neural networks

    Giuseppe Carleo and Matthias Troyer. Solving the quantum many-body problem with artificial neural networks. Science, 355(6325):602–606, 2017

    2017

  21. [62]

    Learning phase transitions by confusion.Nature Physics, 13(5):435–439, 2017

    Evert PL Van Nieuwenburg, Ye-Hua Liu, and Sebastian D Huber. Learning phase transitions by confusion.Nature Physics, 13(5):435–439, 2017

    2017

  22. [63]

    Quantum convolutional neural networks

    Iris Cong, Soonwon Choi, and Mikhail D Lukin. Quantum convolutional neural networks. Nature Physics, 15(12):1273– 1278, 2019

    2019

  23. [64]

    Towards understanding the power of quantum kernels in the nisq era

    Xinbiao Wang, Yuxuan Du, Yong Luo, and Dacheng Tao. Towards understanding the power of quantum kernels in the nisq era. Quantum, 5:531, 2021

    2021

  24. [65]

    The inductive bias of quantum kernels

    Jonas K¨ ubler, Simon Buchholz, and Bernhard Sch¨ olkopf. The inductive bias of quantum kernels. Advances in Neural Information Processing Systems, 34:12661–12673, 2021

    2021

  25. [66]

    Integration with respect to the haar measure on unitary, orthogonal and symplectic group

    Beno ˆ ıt Collins and Piotr´Sniady. Integration with respect to the haar measure on unitary, orthogonal and symplectic group. Communications in Mathematical Physics, 264(3):773–795, 2006

    2006

  26. [67]

    Moments and cumulants of polynomial random variables on unitarygroups, the itzykson-zuber integral, and free probability

    Beno ˆ ıt Collins. Moments and cumulants of polynomial random variables on unitarygroups, the itzykson-zuber integral, and free probability. International Mathematics Research Notices, 2003(17):953–982, 2003

    2003

  27. [68]

    Local random quantum circuits are approximate polynomial-designs

    Fernando GSL Brandao, Aram W Harrow, and Micha l Horodecki. Local random quantum circuits are approximate polynomial-designs. Communications in Mathematical Physics, 346:397–434, 2016

    2016

  28. [69]

    Graham, Donald E

    Ronald L. Graham, Donald E. Knuth, and Oren Patashnik. Concrete Mathematics: A Foundation for Computer Science. Advanced Book Program, Reading, MA, USA, 1st edition, 1989. 10

    1989

  29. [70]

    Flammia, Stephen Becker, and Jens Eisert

    David Gross, Yi-Kai Liu, Steven T. Flammia, Stephen Becker, and Jens Eisert. Quantum state tomography via compressed sensing. Physical Review Letters, 105:150401, Oct 2010

    2010

  30. [71]

    Experimental quantum compressed sensing for a seven-qubit system

    Carlos A Riofrio, David Gross, Steven T Flammia, Thomas Monz, Daniel Nigg, Rainer Blatt, and Jens Eisert. Experimental quantum compressed sensing for a seven-qubit system. Nature communications, 8(1):15305, 2017

    2017

  31. [72]

    Senrui Chen, Wenjun Yu, Pei Zeng, and Steven T. Flammia. Robust shadow estimation. PRX Quantum, 2:030348, Sep 2021

    2021

  32. [73]

    Mitigating measurement errors in multiqubit experiments

    Sergey Bravyi, Sarah Sheldon, Abhinav Kandala, David C Mckay, and Jay M Gambetta. Mitigating measurement errors in multiqubit experiments. Physical Review A, 103(4):042605, 2021

    2021

  33. [74]

    Dephasing-assisted diffusive dynamics in superconducting quantum circuits, 2024

    Yongqi Liang et al. Dephasing-assisted diffusive dynamics in superconducting quantum circuits, 2024

    2024

  34. [75]

    Randomized benchmarking of quantum gates

    Emanuel Knill, Dietrich Leibfried, Rolf Reichle, Joe Britton, R Brad Blakestad, John D Jost, Chris Langer, Roee Ozeri, Signe Seidelin, and David J Wineland. Randomized benchmarking of quantum gates. Physical Review A—Atomic, Molecular, and Optical Physics, 77(1):012307, 2008

    2008

  35. [76]

    Characterizing quantum supremacy in near-term devices

    Sergio Boixo et al. Characterizing quantum supremacy in near-term devices. Nature Physics, 14(6):595–600, 2018

    2018

  36. [77]

    Nature, 614(7949):676–681, 2023

    Suppressing quantum errors by scaling a surface code logical qubit. Nature, 614(7949):676–681, 2023

    2023

  37. [78]

    Experimental quantum adversarial learning with programmable superconducting qubits

    Wenhui Ren et al. Experimental quantum adversarial learning with programmable superconducting qubits. Nature Computational Science, 2(11):711–717, 2022

    2022

  38. [79]

    Dynamical decoupling for superconducting qubits: A performance survey

    Nic Ezzell, Bibek Pokharel, Lina Tewala, Gregory Quiroz, and Daniel A Lidar. Dynamical decoupling for superconducting qubits: A performance survey. Physical Review Applied, 20(6):064027, 2023

    2023

  39. [80]

    Overlapped grouping measurement: A unified framework for measuring quantum states

    Bujiao Wu, Jinzhao Sun, Qi Huang, and Xiao Yuan. Overlapped grouping measurement: A unified framework for measuring quantum states. Quantum, 7:896, 2023

    2023

  40. [81]

    10-qubit entanglement and parallel logic operations with a superconducting circuit

    Chao Song, Kai Xu, Wuxin Liu, Chui-ping Yang, Shi-Biao Zheng, Hui Deng, Qiwei Xie, Keqiang Huang, Qiujiang Guo, Libo Zhang, et al. 10-qubit entanglement and parallel logic operations with a superconducting circuit. Physical review letters, 119(18):180511, 2017

    2017

  41. [82]

    Efficient z gates for quantum computing

    David C McKay, Christopher J Wood, Sarah Sheldon, Jerry M Chow, and Jay M Gambetta. Efficient z gates for quantum computing. Physical Review A, 96(2):022330, 2017

    2017

  42. [83]

    State Similarity in Modular Superconducting Quantum Processors with Classical Communications

    Brooks Foxen, Charles Neill, Andrew Dunsworth, Pedram Roushan, Ben Chiaro, Anthony Megrant, Julian Kelly, Zijun Chen, Kevin Satzinger, Rami Barends, et al. Demonstrating a continuous set of two-qubit gates for near-term quantum algorithms. Physical Review Letters, 125(12):1205...

    2020

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.