betaFeedback

WorksPhysics and Astronomy

Certifying quantum generative models beyond the Fisher trace

Abstract

Whether a trained quantum generative model has learned its target is usually decided by a diagnostic validated on a labeled benchmark, and such a diagnostic can pass that validation by recovering how the benchmark was built. We analyze Fisher-trace and kernel-based certification of the real-amplitude quantum circuit Born machine in noiseless statevector simulation. For real-amplitude circuits built from half-angle rotations, the classical Fisher information equals the quantum Fisher information almost everywhere, so its trace equals the parameter count and carries no information about fit quality. The Fisher eigenvalue rank counts the live nodes of the amplitude tree: it coincides with the label on a sparse benchmark and is uninformative on a full-support one. The pooled Wasserstein and maximum-mean-discrepancy scores move with optimizer convergence and with the scoring convention. For two bitstring defect families on which both metrics have closed forms, we compute the tolerance-mismatch window in which the discrepancy rejects generators that the transport threshold admits; it opens only at certification-grade tolerances and closes as the qubit number grows. On the Hamming cost we prove a sharp tolerance ratio above which the discrepancy vote is redundant, and we obtain the exact worst-parity minimax value over all unit-diagonal positive-semidefinite kernels, which is exponentially small in the qubit number. In a numerical campaign at the tolerances a near-term experiment can resolve, the discrepancy vote changed no verdict at a thousand shots or more. The false-pass guarantee does not extend to device noise.

Introduction

Quantum generative models approximate a target probability distribution by optimizing the parameters of a variational quantum circuit, with applications ranging from quantum-chemistry data generation to financial modeling and combinatorial sampling. Three model families dominate this effort, namely quantum generative adversarial networks (QGANs) [1, 2], which have been used to learn and load random distributions onto quantum hardware, quantum circuit Born machines (QCBMs) [3, 4], and quantum generative diffusion models (QGDMs) [5]. In our own earlier work we used the QCBM as a data-driven benchmark for training shallow circuits on near-term devices [4], and we built an optimized adversarial framework, Re-QGAN, around a real-amplitude generator circuit [6]. As such models move from simulation toward deployment, whether a trained generator has actually learned its target becomes a question of certification, which has to be settled before its samples are used downstream.

In practice, a practitioner certifies a trained generator by scoring its output distribution against the target with a distributional metric, most commonly the Wasserstein-1 distance (W1_1) [1] or the maximum mean discrepancy (MMD) [7]. Channels drawn from the Fisher-information family, whose spectra and effective dimensions were introduced as capacity measures [8, 9, 10], are candidate companions to such scores. Each such diagnostic is trusted once it scores well on a labeled benchmark, returning high values for good generators and low values for defective ones. A benchmark, however, is built by varying a known control, typically a perturbation applied to a good target, so that a diagnostic can score well by recovering that control variable instead of the distributional property it claims to measure. Such a diagnostic reports how the benchmark was constructed, or how the generator was trained, and says nothing about whether the output matches the target.

For the Fisher-information channel, whether it reports fit quality is settled by two known facts: on real-amplitude pure states the classical and quantum Fisher information coincide, a result due to Facchi et al. [11], and a half-angle rotation with an involutive generator contributes a unit entry to the quantum Fisher diagonal [12]. We apply them to the universal real-amplitude state-preparation ansatz of Möttönen et al. [13, 14], used by the QCBM throughout this work and as the Re-QGAN generator [6]: the classical Fisher information matrix of the computational-basis distribution coincides with the quantum Fisher information matrix, and its trace equals the parameter count d=2n−1d = 2^n - 1, at every parameter point outside a measure-zero set on which some basis amplitude vanishes (Proposition 1). The identity extends to all half-angle real-amplitude circuits, including the layered hardware-efficient family. As a consequence, the Fisher trace is constant on the operational parameter space, so that it carries no information about fit quality, and any certification protocol that uses it reduces to its remaining channels.

In this work, we take this identity as the starting observation and ask which channels can still certify a real-amplitude QCBM. We first specialize the identity to the deployed ansatz class, describe the exceptional set explicitly for the Möttönen circuit, and validate the identity on our training campaign. We then test the two remaining candidates, one spectral and one distributional, beginning with the eigenvalue rank of the same Fisher matrix, whose capacity interpretation has been developed for quantum neural networks and for QCBMs [10, 15, 16], and which separates perturbed from unperturbed runs perfectly on the principal benchmark through n=7n = 7. The rank, however, counts the live nodes of the amplitude tree at the resolution of its eigenvalue threshold, which is the rank of the quantum Fisher matrix; the support-restricted classical matrix would have rank at most one. On the two-peak benchmark it equals 2n−12^n - 1 whenever the fit is exact. A severity split, a pre-registered stress test on full-support defect families and a pre-registered perturbed full-support benchmark collapse it to chance, and the pattern recurs under three further optimizers. We keep it only as a flag for anisotropic verdict robustness, a role that a direct perturbation test supports. In turn, the pooled W1W_1 and MMD scores move with optimizer convergence and with the reference distribution against which they are computed; on the sparse benchmark they become perfect once the negative class is conditioned on convergence, and on the full-support benchmark they resolve a defect to the extent that it exceeds the fit residual. For the distributional channels that remain, we analyze an escalation rule that consults MMD only when W1W_1 passes narrowly, and we prove a finite-sample bound on its false-pass probability that improves quadratically on the single-metric rule at the same threshold for defects near it. When it does consult MMD, the rule has a higher false-fail rate than the single-metric rule. For two families of bitstring defects on which both W1W_1 and MMD have closed forms, we compute exactly the tolerance-mismatch region, the set of generators that the transport threshold admits and the discrepancy threshold rejects, together with the window of scales and tolerance ratios in which it is non-empty. The window requires certification-grade tolerances and narrows as the qubit number increases. Inside it the escalation acts on both families as a single shifted transport threshold, and the two families close it

at tolerance ratios that differ by a fixed factor. Against an unknown defect family at a fixed operating point it remains open only at the smallest qubit number we study. On the Hamming cost we prove two further exact limits of any such protocol. The discrepancy is redundant for every pair of distributions once its tolerance exceeds 2(1−e−γ)τW122(1 - e^{-\gamma})\tau_{W_1}^2. For every normalized kernel, some parity defect of fixed transport size has squared discrepancy at most β2/(2n−1)\beta^2/(2^n - 1), so that no fixed kernel is uniformly sensitive to correlation defects and certification must name the correlation orders it is meant to detect. That kernel discrepancies weight high-order correlations weakly is known. Rudolph et al. [17] draw out its consequence for training gradients, and Kurkin et al. [18] exhibit pairs at unit total variation whose discrepancy is exponentially small for every stationary characteristic kernel with unit diagonal on the bitstring space. Our parity statement covers every unit-diagonal positive-semidefinite kernel, stationary or not, takes an explicit parity family at fixed transport distance as its alternatives, and gives an exact value with an attaining kernel (Sec. V F). In this way the present work complements the empirical multi-metric advocacy of Gili et al. [19] and the alternative-loss exploration of Coyle et al. [20] with a structural account of why the trace channel would fail and with finite-sample bounds and design-recovery controls for the protocol that replaces it. Section II fixes notation, the ansatz, the decision rules and the experimental configuration. We prove and validate the Fisher-trace identity in Sec. III, test the rank and the pooled distributional scores in Sec. IV, and define, bound and test the escalation protocol in Sec. V. Section VI discusses alternatives, noise and limitations, and Sec. VII concludes. The proofs and the concentration machinery are collected in Appendices A and B.

Background and experimental design

We consider quantum generative models whose output is a probability distribution q∈ΔNq \in\Delta_N, the probability simplex over the outcome space X={0,1}n\mathcal{X} = \{0,1\}^n with N=2nN = 2^n elements. The distribution is obtained by measuring an nn-qubit parameterized quantum circuit (PQC) in the computational basis. The target distribution p∈ΔNp \in\Delta_N is either known analytically or estimated from samples. The certification question is then: given pp and qq, is qq an acceptable approximation of pp?

Wasserstein-1 distance

The Wasserstein-1 (earth mover’s) distance between distributions pp and qq on X\mathcal{X}, equipped with a ground-cost matrix C∈R≥0N×NC \in\mathbb{R}_{\ge0}^{N \times N}, is defined as the optimal transport cost:

W1(p,q;C)=min⁡Γ≥0⟨C,Γ⟩s.t.Γ1=p,ΓT1=q(1)W_1(p,q;C) = \min_{\Gamma\ge0} \langle C,\Gamma\rangle\quad\text{s.t.}\quad\Gamma\mathbf{1}=p,\quad\Gamma^{\mathsf{T}}\mathbf{1}=q \tag*{(1)}

where Γ\Gamma is the transport plan (coupling), 1\mathbf{1} is the all-ones vector so that the two constraints fix the row and column marginals of Γ\Gamma to pp and qq, and ⟨⋅,⋅⟩\langle\cdot,\cdot\rangle denotes the Frobenius inner product. Equivalently, by Kantorovich duality,

W1(p,q;C)=max⁡f∑x∈Xf(x)(p(x)−q(x)),(2)W_1(p,q;C)=\max_f\sum_{x\in\mathcal{X}}f(x)(p(x)-q(x)), \tag*{(2)}

where the maximum is over CC-Lipschitz potentials: f(x)−f(y)≤C(x,y)f(x)-f(y)\le C(x,y) for all x,yx,y. We denote throughout by TV⁡(p,q)=12∥p−q∥1=12∑x∣p(x)−q(x)∣\operatorname{TV}(p,q)=\frac{1}{2}\lVert p-q\rVert_1=\frac{1}{2}\sum_x\lvert p(x)-q(x)\rvert the total-variation distance, which enters the closed forms of § 4 D.

The ground cost determines which distributional differences W1W_1 sees, and we use three: the Hamming cost C1(x,y)=dH(x,y)C_1(x,y)=d_H(x,y); a weighted variant C2(x,y)=∑iwi∣xi−yi∣C_2(x,y)=\sum_i w_i\lvert x_i-y_i\rvert with qubit weights wi>0w_i>0; and a circuit-structure cost C3(x,y)=dH(x,y)+αW∑i∣χi(x)−χi(y)∣C_3(x,y)=d_H(x,y)+\alpha_W\sum_i\lvert\chi_i(x)-\chi_i(y)\rvert with nearest-neighbor parities χi(x)=xi⊕xi+1\chi_i(x)=x_i\oplus x_{i+1}. All three are metrics (C3 adds parity pseudometrics to the Hamming metric), so the duality (2) and the centering argument of Proposition 6 apply to each. The maximum cost Dmax⁡=max⁡x,yC(x,y)D_{\max}=\max_{x,y}C(x,y) determines the scale of W1W_1 and enters the sample-complexity bounds (§ 4).

Maximum mean discrepancy

The maximum mean discrepancy (MMD) between pp and qq, with respect to a positive-definite kernel k:X×X→Rk:\mathcal{X}\times\mathcal{X}\to\mathbb{R}, is defined according to [7]

MMD⁡k2(p,q)=Ex,x′∼p[k(x,x′)]+Ey,y′∼q[k(y,y′)]−2Ex∼p,y∼q[k(x,y)].(3)\operatorname{MMD}_k^2(p,q)=\mathbb{E}_{x,x'\sim p}[k(x,x')]+\mathbb{E}_{y,y'\sim q}[k(y,y')]-2\mathbb{E}_{x\sim p,y\sim q}[k(x,y)]. \tag*{(3)}

Equivalently, MMD⁡k(p,q)=∥μp−μq∥Hk\operatorname{MMD}_k(p,q)=\lVert\mu_p-\mu_q\rVert_{\mathcal{H}_k}, where μp=Ex∼p[k(x,⋅)]\mu_p=\mathbb{E}_{x\sim p}[k(x,\cdot)] is the kernel mean embedding in the reproducing kernel Hilbert space (RKHS) Hk\mathcal{H}_k.

We use the radial basis function (RBF) kernel on Hamming distance:

k(x,y)=exp⁡(−γdH(x,y)),γ=1/n.(4)k(x,y)=\exp(-\gamma d_H(x,y)),\qquad\gamma=1/n. \tag*{(4)}

For a characteristic kernel such as the RBF, MMD⁡k(p,q)=0\operatorname{MMD}_k(p,q)=0 if and only if p=qp=q [7]. The kernel bandwidth γ=1/n\gamma=1/n is chosen so that the kernel has non-trivial sensitivity across the full range of Hamming distances at each qubit count.

In the statevector campaign MMD⁡2\operatorname{MMD}^2 is evaluated exactly from the output distributions (Sec. II H). The finite-shot study of § 4 J estimates it from samples with the unbiased UU-statistic of Proposition 8 and reports the plug-in (VV-statistic) estimator on the same draws for comparison; the shot-budget study of Appendix B § 5.8.4 quantifies the difference between the two. In every setting the value is thresholded against a tolerance τMMD\tau_{\mathrm{MMD}}. Throughout this paper τMMD\tau_{\mathrm{MMD}} acts on the squared statistic, as in Algorithm 1: the loose benchmark-comparison tolerance is MMD⁡k2≤10−2\operatorname{MMD}_k^2\le10^{-2} and the certification-grade tolerance is MMD⁡k2≤10−4\operatorname{MMD}_k^2\le10^{-4}, equivalently MMD⁡k≤0.1\operatorname{MMD}_k\le0.1 and MMD⁡k≤0.01\operatorname{MMD}_k\le0.01. The labels “loose” and “certification-grade” refer to τMMD\tau_{\mathrm{MMD}} alone; the transport threshold is τW1=0.05\tau_{W_1}=0.05 in every analysis. We do not use a permutation test, the customary two-sample tool, because its pp-value tests the sharp null p=qp=q. Certification asks whether the discrepancy exceeds an operational tolerance, and a generator can differ from its target at any significance level while sitting well inside any tolerance a practitioner would set. The false-pass guarantee of § 4 B therefore covers the threshold comparison.

Classical Fisher-trace diagnostic

The classical Fisher trace measures the generator’s sensitivity to parameter perturbations at the current operating point. For a PQC with parameters θ∈Rd\theta\in\mathbb{R}^d producing output distribution pθp_\theta over X\mathcal{X}, the classical Fisher information matrix (CFIM) is

Fjk(θ)=∑x∈X1pθ(x)∂pθ(x)∂θj∂pθ(x)∂θk.(5)F_{jk}(\theta)=\sum_{x\in\mathcal{X}}\frac{1}{p_\theta(x)}\frac{\partial p_\theta(x)}{\partial\theta_j}\frac{\partial p_\theta(x)}{\partial\theta_k}. \tag*{(5)}

The diagnostic is the trace of the CFIM:

CFI⁡w(θ)=Tr⁡(F(θ))=∑k=1d∑x∈X1pθ(x)(∂pθ(x)∂θk)2.(6)\operatorname{CFI}_w(\theta)=\operatorname{Tr}(F(\theta))=\sum_{k=1}^{d}\sum_{x\in\mathcal{X}}\frac{1}{p_\theta(x)}\left(\frac{\partial p_\theta(x)}{\partial\theta_k}\right)^2. \tag*{(6)}

Here, the CFIM in Eq. (5) is dominated by the Löwner (positive-semidefinite, PSD) partial order by the quantum Fisher information matrix (QFIM) computed from the full density operator [21, 22], an inequality we exploit in § 2 to derive an analytical cap on CFI⁡w\operatorname{CFI}_w. In the noiseless statevector setting, the Jacobians ∂pθ/∂θk\partial p_\theta/\partial\theta_k are computed exactly by the parameter-shift rule (Sec. II H). The Fisher trace is a property of the generator at its current parameters. Fisher spectra and the effective dimensions built on them have been proposed as expressibility and trainability measures [8, 9, 10]; whether the trace can be a certification channel alongside the distributional ones is what § 2 tests on the Möttönen ansatz used in this work.

The Möttönen real-amplitude state-preparation ansatz

We adopt the universal real-amplitude state-preparation circuit of Möttönen et al. [13, 14], both as the QCBM generator and as the working example for the Fisher-trace identity of § 2. It is also the generator architecture of our earlier work on quantum generative adversarial networks (Re-QGAN) [6]. The circuit is recursive: a length-nn block is constructed from two length-(n−1)(n-1) blocks joined by a single CNOT between the first and last qubit, with the base case providing one or two RYR_Y rotations on a single or two-qubit block. Applying such blocks in sequence on the first kk qubits, k=1,…,nk = 1, \ldots, n, gives the full circuit, whose stage kk is the uniformly controlled rotation of Appendix A 2. The parameter count satisfies the recurrence T(n)=2T(n−1)+1T(n) = 2T(n - 1) + 1 with T(1)=1T(1) = 1, yielding exactly d=2n−1d = 2^n - 1 rotation parameters and 2n−n−12^n - n - 1 CNOT gates for an nn-qubit ansatz. This count matches the dimension of the real-amplitude pure-state sphere SR2n−1⊂R2nS_{\mathbb{R}}^{2^n-1} \subset\mathbb{R}^{2^n}, so the ansatz is universal for real-amplitude states with the minimal parameter count compatible with universality.

References

  1. [1]Christa Zoufal, Aurelien Lucchi, and Stefan Woerner, Quantum generative adversarial networks for learning and loading random distributions, npj Quantum Information 5, 103 (2019).
  2. [2]Samuel A. Stein, Betis Baheri, Daniel Chen, Ying Mao, Qiang Guan, Ang Li, Bo Fang, and Shuai Xu, QuGAN: A quantum state fidelity based generative adversarial network, arXiv preprint arXiv:2010.09036 (2020).DOI
  3. [3]Jin-Guo Liu and Lei Wang, Differentiable learning of quantum circuit born machines, Physical Review A 98, 062324 (2018).DOI
  4. [4]Marcello Benedetti, Delfina Garcia-Pintos, Oscar Perdomo, Vicente Leyton-Ortega, Yunseong Nam, and Alejandro Perdomo-Ortiz, A generative modeling approach for benchmarking and training shallow quantum circuits, npj Quantum Information 5, 45 (2019).DOI
  5. [5]Chuangtao Chen, Xingang Zhao, Nan-Run Zhou, Xiliang He, Jiahao Sun, and Haozhen Situ, Quantum generative diffusion model: A fully quantum-mechanical model for generating quantum state ensemble, arXiv preprint arXiv:2401.07039 (2024).DOI
  6. [6]Sandra Nguemto and Vicente Leyton-Ortega, Re-QGAN: An optimized adversarial quantum circuit learning framework, arXiv preprint arXiv:2208.02165 (2022).arxiv.org/abs/2208.02165
  7. [7]Arthur Gretton, Karsten M. Borgwardt, Malte J. Rasch, Bernhard Schölkopf, and Alexander Smola, A kernel two-sample test, Journal of Machine Learning Research 13, 723 (2012).DOI
  8. [8]Amira Abbas, David Sutter, Christa Zoufal, Aurelien Lucchi, Alessio Figalli, and Stefan Woerner, The power of quantum neural networks, Nature Computational Science 1, 403 (2021).DOI
  9. [9]Johannes Jakob Meyer, Fisher information in noisy intermediate-scale quantum applications, Quantum 5, 539 (2021).
  10. [10]Tobias Haug, Kishor Bharti, and M. S. Kim, Capacity and quantum geometry of parametrized quantum circuits, PRX Quantum 2, 040309 (2021).DOI
  11. [11]Paolo Facchi, Ravi Kulkarni, V. I. Man’ko, Giuseppe Marmo, E. C. G. Sudarshan, and Franco Ventriiglia, Classical and quantum Fisher information in the geometrical formulation of quantum mechanics, Physics Letters A 374, 4801 (2010).DOI
  12. [12]Jing Liu, Haidong Yuan, Xiao-Ming Lu, and Xiaoguang Wang, Quantum Fisher information matrix and multiparameter estimation, Journal of Physics A: Mathematical and Theoretical 53, 023001 (2020).DOI
  13. [13]Mikko Möttönen, Juha J. Vartiainen, Ville Bergholm, and Martti M. Salomaa, Transformation of quantum states using uniformly controlled rotations, Quantum Information & Computation 5, 467 (2005).arxiv.org/abs/quant-ph/0407010
  14. [14]Mikko Möttönen, Juha J. Vartiainen, Ville Bergholm, and Martti M. Salomaa, Quantum circuits for general multiqubit gates, Physical Review Letters 93, 130502 (2004).DOI
  15. [15]Martín Larocca, Nathan Ju, Diego García-Martín, Patrick J. Coles, and M. Cerezo, Theory of over-parametrization in quantum neural networks, Nature Computational Science 3, 542 (2023).
  16. [16]Andrea Delgado, Francisco Rios, and Kathleen E. Hamilton, Identifying overparametrization in quantum circuit Born machines (2023), arXiv:2307.03292.arxiv.org/abs/2307.03292
  17. [17]Manuel S. Rudolph, Sacha Lerch, Supanut Thanasilp, Oriel Kiss, Sofia Vallecorsa, Michele Grossi, and Zoë Holmes, Trainability barriers and opportunities in quantum generative modeling, npj Quantum Information 10, 116 (2024).DOI
  18. [18]Andrii Kurkin, Kevin Shen, Susanne Pielawa, Hao Wang, and Vedran Dunjko, Universality and kernel-adaptive training for classically trained, quantum-deployed generative models (2025), arXiv:2510.08476.arxiv.org/abs/2510.08476
  19. [19]Kaitlin Gili, Marta Mauri, and Alejandro Perdomo-Ortiz, Generalization metrics for practical quantum advantage in generative models, Physical Review Applied 21, 044032 (2024).arxiv.org/abs/2201.08770
  20. [20]Brian Coyle, Daniel Mills, Vincent Danos, and Elham Kashefi, The Born supremacy: Quantum advantage and training of an Ising Born machine, npj Quantum Information 6, 60 (2020).DOI
  21. [21]Dominik Šafránek, Simple expression for the quantum Fisher information matrix, Physical Review A 97, 042322 (2018).DOI
  22. [22]Andreas Rath, Cyril Branciard, Anna Minguzzi, and Benoît Vermersch, Quantum Fisher information from randomized measurements, Physical Review Letters 127, 260501 (2021).DOI
  23. [23]Nikolaus Hansen, The CMA evolution strategy: A tutorial, arXiv preprint arXiv:1604.00772 (2016).arxiv.org/abs/1604.00772
  24. [24]Andrew Arrasmith, M. Cerezo, Piotr Czar­nik, Lukasz Cincio, and Patrick J. Coles, Effect of barren plateaus on gradient-free optimization, Quantum 5, 558 (2021).DOI
  25. [25]Ville Bergholm, Josh Isaac, Maria Schuld, Christian Gogolin, Shahnawaz Ahmed, et al., PennyLane: Automatic differentiation of hybrid quantum-classical computations, arXiv preprint arXiv:1811.04968 (2018).arxiv.org/abs/1811.04968
  26. [26]Carl W. Helstrom, Quantum Detection and Estimation Theory (Academic Press, New York, 1976).
  27. [27]Alexander S. Holevo, Probabilistic and Statistical Aspects of Quantum Theory (North-Holland, Amsterdam, 1982).
  28. [28]Olivier Roy and Martin Vetterli, The effective rank: A measure of effective dimensionality, in Proceedings of the 15th European Signal Processing Conference (EUSIPCO) (2007) pp. 606–610.
  29. [29]Jean Feydy, Thibault Séjourné, François-Xavier Vialard, Shun-ichi Amari, Alain Trouvé, and Gabriel Peyré, Interpolating between optimal transport and MMD using Sinkhorn divergences, in Proceedings of the 22nd International Conference on Artificial Intelligence and Statistics (AISTATS), Proceedings of Machine Learning Research, Vol. 89 (2019) pp. 2681–2690.
  30. [30]Jonathan Weed and Francis Bach, Sharp asymptotic and finite-sample rates of convergence of empirical measures in Wasserstein distance, Bernoulli 25, 2620 (2019).arxiv.org/abs/1707.00087
  31. [31]Roger L. Berger and Jason C. Hsu, Bioequivalence trials, intersection-union tests and equivalence confidence sets, Statistical Science 11, 283 (1996).DOI
  32. [32]Bharath K. Sriperumbudur, Arthur Gretton, Kenji Fukumizu, Bernhard Schölkopf, and Gert R. G. Lanckriet, Hilbert space embeddings and metrics on probability measures, Journal of Machine Learning Research 11, 1517 (2010).arxiv.org/abs/0907.5309
  33. [33]Mikołaj Bińkowski, Dougal J. Sutherland, Michael Arbel, and Arthur Gretton, Demystifying MMD GANs, in International Conference on Learning Representations (ICLR) (2018) arXiv:1801.01401.arxiv.org/abs/1801.01401
  34. [34]Felix Biggs, Antonin Schrab, and Arthur Gretton, MMD-FUSE: Learning and combining kernels for two-sample testing without data splitting, in Advances in Neural Information Processing Systems, Vol. 36 (2023) arXiv:2306.08777.arxiv.org/abs/2306.08777
  35. [35]Titouan Vayer and Rémi Gribonval, Controlling Wasserstein distances by kernel norms with application to compressive statistical learning, Journal of Machine Learning Research 24, 1 (2023).
  36. [36]Thibault Modeste and Clément Dombry, Characterization of translation invariant MMD on R^d and connections with Wasserstein distances, Journal of Machine Learning Research 25, 1 (2024).
  37. [37]Giacomo De Palma, Milad Marvian, Dario Trevisan, and Seth Lloyd, The quantum Wasserstein distance of order 1, IEEE Transactions on Information Theory 67, 6627 (2021).
  38. [38]Emily Beatty and Daniel Stilck França, Order p quantum Wasserstein distances from couplings, arXiv preprint arXiv:2402.16477 (2024).DOI
  39. [39]Vicente Leyton-Ortega and Kathleen Hamilton, Data and code for “Certifying quantum generative models beyond the Fisher trace”, Zenodo (2026), DOI to be assigned on deposit.
  40. [40]Matteo G. A. Paris, Quantum estimation for quantum technology, International Journal of Quantum Information 7, 125 (2009).
  41. [41]Antonio Falcó, Daniela Falcó-Pomares, and Hermann G. Matthies, A rigorous and self-contained proof of the Grover–Rudolph state preparation algorithm (2026), arXiv:2601.17930.arxiv.org/abs/2601.17930
  42. [42]Colin McDiarmid, On the method of bounded differences, in Surveys in Combinatorics, London Mathematical Society Lecture Note Series, Vol. 141, edited by J. Siemons (Cambridge University Press, 1989) pp. 148–188.DOI
  43. [43]Nicolas Fournier and Arnaud Guillin, On the rate of convergence in Wasserstein distance of the empirical measure, Probability Theory and Related Fields 162, 707 (2015).arxiv.org/abs/1312.2128

Paper details

vlnc:W_bvmry4hj
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 5

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents