pith. sign in

arxiv: 2505.11025 · v1 · submitted 2025-05-16 · 🪐 quant-ph · cs.IT· cs.LG· math.IT

Generalization Bounds for Quantum Learning via R\'enyi Divergences

Naqueeb Ahmad Warsi , Ayanava Dasgupta , Masahito Hayashi This is my paper

Pith reviewed 2026-05-22 15:06 UTC · model grok-4.3

classification 🪐 quant-ph cs.ITcs.LGmath.IT
keywords generalization boundsquantum learningRényi divergencesPetz divergencemodified sandwich divergencevariational evaluationquantum machine learningprobabilistic bounds
0
0 comments X

The pith

Quantum learning generalization errors are upper-bounded by Rényi divergences, and a new modified sandwich quantum version gives tighter results than the Petz divergence.

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

The paper derives a family of upper bounds on the expected generalization error of quantum learning algorithms. It extends a prior framework with a new definition of expected true loss and evaluates the bounds through a variational approach to quantum Rényi divergences. Both quantum and classical Rényi divergences appear in the expressions. Analytical and numerical comparisons establish that the modified sandwich quantum Rényi divergence produces superior bounds relative to the Petz divergence. Probabilistic generalization error bounds are also supplied using the modified sandwich divergence together with classical and smooth max Rényi divergences.

Core claim

Upper bounds on the expected generalization error of quantum learning algorithms are obtained in terms of quantum and classical Rényi divergences by a variational evaluation method; the newly introduced modified sandwich quantum Rényi divergence is shown analytically and numerically to yield tighter bounds than the Petz divergence, and probabilistic bounds follow from two further techniques based on the modified sandwich and smooth max Rényi divergences.

What carries the argument

Variational evaluation of the Petz and modified sandwich quantum Rényi divergences to produce upper bounds on expected generalization error.

If this is right

  • Tighter theoretical guarantees become available for predicting how quantum models perform on unseen data.
  • The modified sandwich divergence can replace the Petz divergence in existing analyses to obtain improved error estimates.
  • Probabilistic bounds supply concentration results that quantify how often the generalization error stays close to its expectation.
  • Hybrid bounds mixing quantum and classical Rényi divergences apply directly to mixed classical-quantum learning pipelines.

Where Pith is reading between the lines

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

  • The variational technique could be repurposed to design training objectives that explicitly minimize the relevant Rényi divergence.
  • Similar bounds may extend to variational quantum algorithms or quantum neural networks on near-term hardware.
  • Numerical superiority of the sandwich divergence hints that it better reflects the quantum correlations relevant to generalization.

Load-bearing premise

The framework from Caro et al. together with the new definition of expected true loss remains valid and permits variational evaluation of the quantum Rényi divergences.

What would settle it

Compute the modified sandwich and Petz bounds for a concrete quantum learning task with known true risk and check whether the sandwich bounds are observably tighter than the Petz bounds or whether they fail to upper-bound the actual generalization error.

Figures

Figures reproduced from arXiv: 2505.11025 by Ayanava Dasgupta, Masahito Hayashi, Naqueeb Ahmad Warsi.

Figure 1
Figure 1. Figure 1: Quantum learning algorithm structure proposed by [8]. [PITH_FULL_IMAGE:figures/full_fig_p014_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: RHS of eq. (88) vs RHS of eq. (98) vs RHS of eq. (100). [PITH_FULL_IMAGE:figures/full_fig_p027_2.png] view at source ↗
read the original abstract

This work advances the theoretical understanding of quantum learning by establishing a new family of upper bounds on the expected generalization error of quantum learning algorithms, leveraging the framework introduced by Caro et al. (2024) and a new definition for the expected true loss. Our primary contribution is the derivation of these bounds in terms of quantum and classical R\'enyi divergences, utilizing a variational approach for evaluating quantum R\'enyi divergences, specifically the Petz and a newly introduced modified sandwich quantum R\'enyi divergence. Analytically and numerically, we demonstrate the superior performance of the bounds derived using the modified sandwich quantum R\'enyi divergence compared to those based on the Petz divergence. Furthermore, we provide probabilistic generalization error bounds using two distinct techniques: one based on the modified sandwich quantum R\'enyi divergence and classical R\'enyi divergence, and another employing smooth max R\'enyi divergence.

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

2 major / 2 minor

Summary. The paper derives a new family of upper bounds on the expected generalization error of quantum learning algorithms by extending the Caro et al. (2024) framework with a novel definition of expected true loss. It evaluates these bounds variationally using the Petz quantum Rényi divergence and a newly introduced modified sandwich quantum Rényi divergence, claiming analytically and numerically that the latter yields superior (tighter) bounds. The work also presents two sets of probabilistic generalization error bounds, one combining the modified sandwich quantum Rényi divergence with classical Rényi divergence and another using smooth max Rényi divergence.

Significance. If the derivations are valid, the results advance the theoretical toolkit for analyzing generalization in quantum machine learning by supplying a parameterized family of Rényi-based bounds that appear tighter than prior Petz-based versions. The numerical comparisons and the provision of both expected and high-probability bounds constitute concrete, falsifiable contributions that could inform algorithm design and sample-complexity estimates in quantum settings.

major comments (2)
  1. [Main derivation following the new expected true loss definition (abstract and §3–4)] The central bounds rest on the new definition of expected true loss introduced after invoking the Caro et al. (2024) framework. No explicit verification is provided that this definition reduces to the classical case or preserves the variational representation properties required for the Rényi divergences to upper-bound the true generalization error when quantum channels or states deviate from the implicit assumptions. This compatibility is load-bearing for both the expected-error bounds and the claimed superiority of the modified sandwich divergence.
  2. [Analytical comparisons and numerical demonstrations (abstract and §5)] The analytical and numerical demonstrations that the modified sandwich quantum Rényi divergence outperforms the Petz divergence (abstract) rely on the variational evaluation remaining faithful under the new loss. Without a direct comparison or counter-example check showing that the variational optimum still upper-bounds the generalization error for the modified sandwich case, the superiority claim cannot be fully substantiated.
minor comments (2)
  1. [Preliminaries or definition section] Notation for the modified sandwich quantum Rényi divergence should be introduced with an explicit equation number and contrasted with the standard sandwich and Petz definitions to avoid ambiguity.
  2. [Probabilistic bounds] The probabilistic bounds section would benefit from a short table summarizing the different Rényi orders and the resulting sample-complexity scalings.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their careful reading of the manuscript and for providing constructive feedback. We address each of the major comments in detail below and have revised the manuscript to incorporate clarifications and additional verifications where appropriate.

read point-by-point responses

  1. Referee: [Main derivation following the new expected true loss definition (abstract and §3–4)] The central bounds rest on the new definition of expected true loss introduced after invoking the Caro et al. (2024) framework. No explicit verification is provided that this definition reduces to the classical case or preserves the variational representation properties required for the Rényi divergences to upper-bound the true generalization error when quantum channels or states deviate from the implicit assumptions. This compatibility is load-bearing for both the expected-error bounds and the claimed superiority of the modified sandwich divergence.

    Authors: We thank the referee for pointing out the need for explicit verification of the new expected true loss definition. In the revised manuscript, we have added a dedicated paragraph in Section 3 demonstrating that the definition reduces to the classical expected loss when the quantum states are diagonal in the computational basis, corresponding to the classical limit. Regarding the variational representation properties, we clarify that the derivation of the bounds relies on the variational characterization of the Rényi divergences applied to the expectation of the loss. Since the new definition maintains the structure of an expectation (albeit in a quantum setting), the same variational upper bound applies without requiring additional assumptions beyond those in Caro et al. (2024). We have included this explanation to ensure the compatibility is clear. revision: yes

  2. Referee: [Analytical comparisons and numerical demonstrations (abstract and §5)] The analytical and numerical demonstrations that the modified sandwich quantum Rényi divergence outperforms the Petz divergence (abstract) rely on the variational evaluation remaining faithful under the new loss. Without a direct comparison or counter-example check showing that the variational optimum still upper-bounds the generalization error for the modified sandwich case, the superiority claim cannot be fully substantiated.

    Authors: The analytical demonstrations in Section 5 compare the two divergences directly through their definitions and established inequalities, showing that the modified sandwich version yields smaller values in relevant regimes, leading to tighter bounds. The numerical results are obtained by performing the variational optimization for each divergence under the new loss definition. To address the referee's concern about explicit confirmation, we have added in the revision a direct numerical comparison of the variational optima and a simple analytical example where we verify that the bound holds and is tighter for the modified sandwich divergence. This substantiates that the variational evaluation remains valid and superior under the new loss. revision: yes

Circularity Check

0 steps flagged

Derivation builds on external framework with independent new loss definition and variational bounds

full rationale

The paper explicitly starts from the Caro et al. (2024) framework as an external reference and introduces a new definition for the expected true loss before deriving the family of upper bounds on generalization error via quantum and classical Rényi divergences. The variational evaluation of the Petz and modified sandwich divergences, along with the analytical and numerical comparison of their performance, constitutes independent content that does not reduce to the inputs by construction. No self-citation load-bearing step, fitted-input prediction, or self-definitional reduction is present in the described chain; the central claims remain self-contained against the cited external framework and the newly introduced elements.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 1 invented entities

The central claims rest on the prior Caro et al. framework and on the validity of the newly introduced modified sandwich divergence definition.

axioms (1)
  • domain assumption Framework introduced by Caro et al. (2024) for generalization bounds in quantum learning
    Explicitly leveraged as the base for the new bounds.
invented entities (1)
  • modified sandwich quantum Rényi divergence no independent evidence
    purpose: To obtain tighter generalization bounds than those from the Petz divergence
    Newly introduced in this work for the purpose of improving bound performance.

pith-pipeline@v0.9.0 · 5697 in / 1229 out tokens · 47082 ms · 2026-05-22T15:06:29.514395+00:00 · methodology

discussion (0)

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

Lean theorems connected to this paper

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

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

52 extracted references · 52 canonical work pages · 3 internal anchors

  1. [1]

    Lower bounds on learning pauli channels with individual measurements,

    O. Fawzi, A. Oufkir, and D. S. França, “Lower bounds on learning pauli channels with individual measurements,” IEEE Transactions on Information Theory , vol. 71, no. 4, pp. 2642–2661, 2025

    work page 2025

  2. [2]

    Optimal universal learning machines for quantum state discrimination,

    M. Fanizza, A. Mari, and V . Giovannetti, “Optimal universal learning machines for quantum state discrimination,” IEEE Transactions on Information Theory , vol. 65, no. 9, pp. 5931–5944, 2019

    work page 2019

  3. [3]

    Quantum differentially private sparse regression learning,

    Y . Du, M.-H. Hsieh, T. Liu, S. You, and D. Tao, “Quantum differentially private sparse regression learning,” IEEE Transactions on Information Theory , vol. 68, no. 8, pp. 5217–5233, 2022

    work page 2022

  4. [4]

    The quantum wasserstein distance of order 1,

    G. De Palma, M. Marvian, D. Trevisan, and S. Lloyd, “The quantum wasserstein distance of order 1,” IEEE Transactions on Information Theory , vol. 67, no. 10, pp. 6627–6643, 2021

    work page 2021

  5. [5]

    Quantum differential privacy in the local model,

    A. Angrisani and E. Kashefi, “Quantum differential privacy in the local model,” IEEE Transactions on Information Theory , vol. 71, no. 5, pp. 3675–3692, 2025

    work page 2025

  6. [6]

    Limit distribution theory for quantum divergences,

    S. Sreekumar and M. Berta, “Limit distribution theory for quantum divergences,” IEEE Transactions on Information Theory , vol. 71, no. 1, pp. 459–484, 2025

    work page 2025

  7. [7]

    Quantum differential privacy: An information theory perspective,

    C. Hirche, C. Rouzé, and D. S. França, “Quantum differential privacy: An information theory perspective,” IEEE Transactions on Information Theory , vol. 69, no. 9, pp. 5771–5787, 2023

    work page 2023

  8. [8]

    Information-theoretic generalization bounds for learning from quantum data,

    M. C. Caro, T. Gur, C. Rouzé, D. Stilck França, and S. Subramanian, “Information-theoretic generalization bounds for learning from quantum data,” in Proceedings of Thirty Seventh Conference on Learning Theory , ser. Proceedings of Machine Learning Research, S. Agrawal and A. Roth, Eds., vol. 247. PMLR, 30 Jun–03 Jul 2024, pp. 775–839. [Online]. Available:...

    work page 2024

  9. [9]

    On the uniform convergence of relative frequencies of events to their probabilities,

    V . N. Vapnik and A. Y . Chervonenkis, “On the uniform convergence of relative frequencies of events to their probabilities,” Theory of Probability & Its Applications, vol. 16, no. 2, pp. 264–280, 1971. [Online]. Available: https://doi.org/10.1137/1116025

    work page doi:10.1137/1116025 1971

  10. [10]

    A theory of the learnable,

    L. G. Valiant, “A theory of the learnable,” Commun. ACM, vol. 27, no. 11, p. 1134–1142, Nov. 1984. [Online]. Available: https://doi.org/10.1145/1968.1972

    work page doi:10.1145/1968.1972 1984

  11. [11]

    Xu and B.-y

    D. A. McAllester, “Some pac-bayesian theorems,” Machine Learning , vol. 37, no. 3, pp. 355–363, Dec 1999. [Online]. Available: https://doi.org/10.1023/A: 1007618624809

    work page doi:10.1023/a: 1999

  12. [12]

    Information-theoretic upper and lower bounds for statistical estimation,

    T. Zhang, “Information-theoretic upper and lower bounds for statistical estimation,” IEEE Transactions on Information Theory , vol. 52, no. 4, pp. 1307–1321, 2006

    work page 2006

  13. [13]

    Generalization bounds: Perspectives from information theory and pac-bayes,

    F. Hellström, G. Durisi, B. Guedj, and M. Raginsky, “Generalization bounds: Perspectives from information theory and pac-bayes,” 2024. [Online]. Available: https://arxiv.org/abs/2309.04381

    work page arXiv 2024

  14. [14]

    A PAC-Bayesian Tutorial with A Dropout Bound

    D. McAllester, “A pac-bayesian tutorial with a dropout bound,” 2013. [Online]. Available: https://arxiv.org/abs/1307.2118

    work page internal anchor Pith review Pith/arXiv arXiv 2013

  15. [15]

    Controlling bias in adaptive data analysis using information theory,

    D. Russo and J. Zou, “Controlling bias in adaptive data analysis using information theory,” in Proceedings of the 19th International Conference on Artificial Intelligence and Statistics , ser. Proceedings of Machine Learning Research, A. Gretton and C. C. Robert, Eds., vol. 51. Cadiz, Spain: PMLR, 09–11 May 2016, pp. 1232–1240. [Online]. Available: https:...

    work page 2016

  16. [16]

    Information-theoretic analysis of generalization capability of learning algorithms,

    A. Xu and M. Raginsky, “Information-theoretic analysis of generalization capability of learning algorithms,” in Advances in Neural Information Processing Systems, I. Guyon, U. V . Luxburg, S. Bengio, H. Wallach, R. Fergus, S. Vishwanathan, and R. Garnett, Eds., vol. 30. Curran Associates, Inc., 2017. [Online]. Available: https://proceedings.neurips.cc/pap...

    work page 2017

  17. [17]

    Tightening mutual information-based bounds on generalization error,

    Y . Bu, S. Zou, and V . V . Veeravalli, “Tightening mutual information-based bounds on generalization error,” IEEE Journal on Selected Areas in Information Theory, vol. 1, no. 1, p. 121–130, May 2020. [Online]. Available: http://dx.doi.org/10.1109/JSAIT.2020.2991139

    work page doi:10.1109/jsait.2020.2991139 2020

  18. [18]

    Information-theoretic generalization bounds for sgld via data-dependent estimates,

    J. Negrea, M. Haghifam, G. K. Dziugaite, A. Khisti, and D. M. Roy, “Information-theoretic generalization bounds for sgld via data-dependent estimates,” Advances in Neural Information Processing Systems , vol. 32, 2019

    work page 2019

  19. [19]

    Sharpened generalization bounds based on conditional mutual information and an application to noisy, iterative algorithms,

    M. Haghifam, J. Negrea, A. Khisti, D. M. Roy, and G. K. Dziugaite, “Sharpened generalization bounds based on conditional mutual information and an application to noisy, iterative algorithms,” Advances in Neural Information Processing Systems , vol. 33, pp. 9925–9935, 2020

    work page 2020

  20. [20]

    A new family of generalization bounds using samplewise evaluated cmi,

    F. Hellström and G. Durisi, “A new family of generalization bounds using samplewise evaluated cmi,” Advances in Neural Information Processing Systems , vol. 35, pp. 10 108–10 121, 2022

    work page 2022

  21. [21]

    Rényi divergence based bounds on generalization error,

    E. Modak, H. Asnani, and V . M. Prabhakaran, “Rényi divergence based bounds on generalization error,” in 2021 IEEE Information Theory Workshop (ITW) , 2021, pp. 1–6

    work page 2021

  22. [22]

    Rate-distortion theoretic generalization bounds for stochastic learning algorithms,

    M. Sefidgaran, A. Gohari, G. Richard, and U. Simsekli, “Rate-distortion theoretic generalization bounds for stochastic learning algorithms,” in Proceedings of Thirty Fifth Conference on Learning Theory , ser. Proceedings of Machine Learning Research, P.-L. Loh and M. Raginsky, Eds., vol. 178. PMLR, 02–05 Jul 2022, pp. 4416–4463. [Online]. Available: https...

    work page 2022

  23. [23]

    Rate-distortion theoretic bounds on generalization error for distributed learning,

    M. Sefidgaran, R. Chor, and A. Zaidi, “Rate-distortion theoretic bounds on generalization error for distributed learning,” Advances in Neural Information Processing Systems, vol. 35, pp. 19 687–19 702, 2022

    work page 2022

  24. [24]

    Pac-bayesian supervised classification: The thermodynamics of statistical learning,

    O. Catoni, “Pac-bayesian supervised classification: The thermodynamics of statistical learning,” Lecture Notes-Monograph Series , vol. 56, pp. i–163, 2007. [Online]. Available: http://www.jstor.org/stable/20461499

    work page arXiv 2007

  25. [25]

    Generalization error bounds via rényi-, f-divergences and maximal leakage,

    A. R. Esposito, M. Gastpar, and I. Issa, “Generalization error bounds via rényi-, f-divergences and maximal leakage,” IEEE Transactions on Information Theory, vol. 67, no. 8, pp. 4986–5004, 2021

    work page 2021

  26. [26]

    Asymptotic evaluation of certain markov process expectations for large time, i,

    M. D. Donsker and S. R. S. Varadhan, “Asymptotic evaluation of certain markov process expectations for large time, i,” Communications on Pure and Applied Mathematics, vol. 28, no. 1, pp. 1–47, 1975. [Online]. Available: https://onlinelibrary.wiley.com/doi/abs/10.1002/cpa.3160280102 32

    work page doi:10.1002/cpa.3160280102 1975

  27. [27]

    Freed, Michael J

    S. Boucheron, G. Lugosi, and P. Massart, Concentration Inequalities: A Nonasymptotic Theory of Independence . Oxford University Press, 02 2013. [Online]. Available: https://doi.org/10.1093/acprof:oso/9780199535255.001.0001

    work page doi:10.1093/acprof:oso/9780199535255.001.0001 2013

  28. [28]

    Variational representations and neural network estimation of rényi divergences,

    J. Birrell, P. Dupuis, M. A. Katsoulakis, L. Rey-Bellet, and J. Wang, “Variational representations and neural network estimation of rényi divergences,” SIAM Journal on Mathematics of Data Science , vol. 3, no. 4, pp. 1093–1116, 2021. [Online]. Available: https://doi.org/10.1137/20M1368926

    work page doi:10.1137/20m1368926 2021

  29. [29]

    Quasi-entropies for finite quantum systems,

    D. Petz, “Quasi-entropies for finite quantum systems,” Reports on Mathematical Physics , vol. 23, no. 1, pp. 57–65, 1986. [Online]. Available: https://www.sciencedirect.com/science/article/pii/0034487786900674

    work page arXiv 1986

  30. [30]

    On quan- tum Rényi entropies: A new generalization and some properti es

    M. Müller-Lennert, F. Dupuis, O. Szehr, S. Fehr, and M. Tomamichel, “On quantum rényi entropies: A new generalization and some properties,” Journal of Mathematical Physics, vol. 54, no. 12, p. 122203, 12 2013. [Online]. Available: https://doi.org/10.1063/1.4838856

    work page doi:10.1063/1.4838856 2013

  31. [31]

    Wilde, Andreas Winter, and Dong Yang

    M. M. Wilde, A. Winter, and D. Yang, “Strong converse for the classical capacity of entanglement-breaking and hadamard channels via a sandwiched rényi relative entropy,” Communications in Mathematical Physics , vol. 331, no. 2, pp. 593–622, Oct 2014. [Online]. Available: https://doi.org/10.1007/s00220-014-2122-x

    work page doi:10.1007/s00220-014-2122-x 2014

  32. [32]

    On measures of entropy and information,

    A. Rényi, “On measures of entropy and information,” Proceeding of the Fourth Berkeley Symposium on Mathematical Statistics and Probabiltiy , vol. 1, pp. 547–561, 1961

    work page 1961

  33. [33]

    Simple one-shot bounds for various source coding problems using smooth rényi quantities,

    N. A. Warsi, “Simple one-shot bounds for various source coding problems using smooth rényi quantities,” Problems of Information Transmission , vol. 52, no. 1, pp. 39–65, Jan 2016. [Online]. Available: https://doi.org/10.1134/S0032946016010051

    work page doi:10.1134/s0032946016010051 2016

  34. [34]

    Fenchel, On conjugate convex functions

    W. Fenchel, On conjugate convex functions . Springer, 2014

    work page 2014

  35. [35]

    Bhatia, Matrix analysis

    R. Bhatia, Matrix analysis. Springer Science & Business Media, 2013, vol. 169

    work page 2013

  36. [36]

    Distinguishability and Accessible Information in Quantum Theory

    C. A. Fuchs, “Distinguishability and accessible information in quantum theory,” 1996. [Online]. Available: https://arxiv.org/abs/quant-ph/9601020

    work page internal anchor Pith review Pith/arXiv arXiv 1996

  37. [37]

    Sur les fonctions convexes et les inégalités entre les valeurs moyennes,

    J. L. W. V . Jensen, “Sur les fonctions convexes et les inégalités entre les valeurs moyennes,” Acta Mathematica , vol. 30, no. 1, pp. 175–193, Dec 1906. [Online]. Available: https://doi.org/10.1007/BF02418571

    work page doi:10.1007/bf02418571 1906

  38. [38]

    M. J. Wainwright, High-Dimensional Statistics: A Non-Asymptotic Viewpoint , ser. Cambridge Series in Statistical and Probabilistic Mathematics. Cambridge University Press, 2019

    work page 2019

  39. [39]

    The golden-thompson trace inequality is complemented,

    F. Hiai and D. Petz, “The golden-thompson trace inequality is complemented,” Linear Algebra and its Applications , vol. 181, pp. 153–185, 1993. [Online]. Available: https://www.sciencedirect.com/science/article/pii/002437959390029N

    work page arXiv 1993

  40. [40]

    Hayashi, Quantum Information Theory

    M. Hayashi, Quantum Information Theory . United States: Springer Cham, 2017

    work page 2017

  41. [41]

    On variational expressions for quantum relative entropies,

    M. Berta, O. Fawzi, and M. Tomamichel, “On variational expressions for quantum relative entropies,” Letters in Mathematical Physics , vol. 107, no. 12, pp. 2239–2265, Dec 2017. [Online]. Available: https://doi.org/10.1007/s11005-017-0990-7

    work page doi:10.1007/s11005-017-0990-7 2017

  42. [42]

    Frank and Elliott H

    R. L. Frank and E. H. Lieb, “Monotonicity of a relative rényi entropy,” Journal of Mathematical Physics , vol. 54, no. 12, p. 122201, 12 2013. [Online]. Available: https://doi.org/10.1063/1.4838835

    work page doi:10.1063/1.4838835 2013

  43. [43]

    Variational expressions and uhlmann theorem for measured divergences,

    K. Fang, H. Fawzi, and O. Fawzi, “Variational expressions and uhlmann theorem for measured divergences,” 2025. [Online]. Available: https://arxiv.org/abs/2502.07745

    work page arXiv 2025

  44. [44]

    On an inequality of lieb and thirring,

    H. Araki, “On an inequality of lieb and thirring,” Letters in Mathematical Physics , vol. 19, no. 2, pp. 167–170, Feb 1990. [Online]. Available: https://doi.org/10.1007/BF01045887

    work page doi:10.1007/bf01045887 1990

  45. [45]

    Lieb and W

    E. Lieb and W. Thirring, Inequalities for the moments of the eigenvalues of the schrödinger hamiltonian and their relation to sobolev inequalities . Springer Berlin Heidelberg, 2005, pp. 205–239

    work page 2005

  46. [46]

    M. M. Wilde, Quantum Information Theory . Cambridge University Press, 2013. [Online]. Available: https://doi.org/10.1017/CBO9781139525343

    work page doi:10.1017/cbo9781139525343 2013

  47. [47]

    Generalization bounds via information density and conditional information density,

    F. Hellström and G. Durisi, “Generalization bounds via information density and conditional information density,” IEEE Journal on Selected Areas in Information Theory, vol. 1, no. 3, pp. 824–839, 2020

    work page 2020

  48. [48]

    Generalized quantum stein’s lemma and second law of quantum resource theories,

    M. Hayashi and H. Yamasaki, “Generalized quantum stein’s lemma and second law of quantum resource theories,” 2024, arXiv:2408.02722

    work page arXiv 2024

  49. [49]

    One-shot private classical capacity of quantum wiretap channel: Based on one-shot quantum covering lemma,

    J. Radhakrishnan, P. Sen, and N. A. Warsi, “One-shot private classical capacity of quantum wiretap channel: Based on one-shot quantum covering lemma,”

    work page

  50. [50]

    One-Shot Private Classical Capacity of Quantum Wiretap Channel: Based on one-shot quantum covering lemma

    [Online]. Available: https://arxiv.org/abs/1703.01932

    work page internal anchor Pith review Pith/arXiv arXiv

  51. [51]

    A one-shot achievability result for quantum state redistribution,

    A. Anshu, R. Jain, and N. A. Warsi, “A one-shot achievability result for quantum state redistribution,” IEEE Transactions on Information Theory , vol. 64, no. 3, pp. 1425–1435, 2018

    work page 2018

  52. [52]

    Building blocks for communication over noisy quantum networks,

    ——, “Building blocks for communication over noisy quantum networks,” IEEE Transactions on Information Theory , vol. 65, no. 2, pp. 1287–1306, 2019. APPENDIX A PROOF OF LEMMA 1 We define L′ := L − Tr[Lρ]I and note that Tr[L′ρ] = 0. Let L′ and ρ have the following eigen decomposition, L′ = |H|X i=1 αi|i⟩⟨i|, where ∀i ∈ [|H|], a − Tr[Lρ] < α i < b − Tr[Lρ], ...

    work page 2019