Operator convexity along lines, self-concordance, and sandwiched R\'enyi entropies

Hamza Fawzi; James Saunderson; Kerry He

Operator convexity along lines, self-concordance, and sandwiched R\'enyi entropies

T0 review · 0 major / 5 minor · reviewed 2026-08-08 · deepseek-v4-flash

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

Pith's one-line read A one-line convexity test yields optimal self-concordant barriers for sandwiched Rényi entropies.

desk verdict Optimal self-concordant barriers for sandwiched Rényi entropy cones are constructed via a clean generalization of compatibility; the main caveat is a load-bearing cited lemma from Hiai that a referee should ask to be stated precisely. read the letter →

arxiv 2502.05627 v1 pith:ERAN6SGM submitted 2025-02-08 math.OC quant-ph

Kerry He , James Saunderson , Hamza Fawzi This is my paper · ORCID

classification math.OCquant-ph MSC 90C2590C5147A6381P45

keywords self-concordantbarrieroperatorconvexityconcavityalonglinessandwichedRényientropyquantumrelativeinterior-pointmethodslogarithmicallyhomogeneouscompatibility

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 establishes a short bridge between operator convexity and convex optimization: if a sufficiently smooth cone-valued concave function is operator concave along every one-dimensional slice of its domain, then the natural logarithmic barrier of its hypograph is self-concordant. The proof runs through Nesterov and Nemirovskii's compatibility concept, and the one-dimensional condition supplies the required third-derivative bound with parameter one. The main application is the sandwiched Rényi quasi-relative entropy from quantum information: for $\alpha\in[1/2,1]$ the logarithmically homogeneous barrier for the hypograph is self-concordant with the optimal parameter $1+2n$, and for $\alpha\in[1,2]$ the same holds for the epigraph. These are the first self-concordant barriers known for this family, making the corresponding entropy optimization problems accessible to interior-point methods.

What carries the argument

The load-bearing object is the scalar line restriction $F(t)=\langle z,f(x+th)\rangle$. Operator concavity of $F$ on $(-1,1)$ gives the integral representation $F(t)=F(0)+F'(0)t+\tfrac12 F''(0)\int_{-1}^{1}\frac{t^2}{1-st}\,d\mu(s)$ for a probability measure $\mu$. The kernel $\xi_s(t)=t^2/(1-st)$ satisfies $\xi_s'''(0)=6s\ge -6=-3\xi_s''(0)$ for every $s\in[-1,1]$; combined with $F''(0)\le 0$, this yields $\langle z,D^3f(x)[h,h,h]\rangle\le -3\langle z,D^2f(x)[h,h]\rangle$ for every $z$, which is $(K,1)$-compatibility. For the Rényi application, the proof that $F$ is operator concave for $\alpha\in[1/2,1]$ goes through the transposed function $\hat F(t)=tF(1/t)$ and a quoted analytic-continuation result identifying it as a Pick function; the $\alpha\in[1,2]$ case is assembled from noncommutative perspectives of operator concave functions using direct sums and unitary covariance.

What would settle it

Take $n=2$, $\alpha=3/4$, $X=Y=I$, $H=\operatorname{diag}(1/2,-1/2)$, $V=0$, and evaluate $\hat F(z)=\operatorname{tr}[((zI)^{(1-\alpha)/(2\alpha)}(zI+H)(zI)^{(1-\alpha)/(2\alpha)})^\alpha]$ for $z$ with positive imaginary part; if the imaginary part of $\hat F(z)$ is ever nonpositive, Lemma 4.2 is false and the paper's hypograph barrier for $\alpha\in[1/2,1]$ no longer follows. A numerical scan over many boundary-feasible $H,V$ and $\alpha$ values could settle whether the quoted continuation holds on the exact stated domain.

Watch

Extended reading notes

Core claim

The central claim is Theorem 3.1: let $f$ be a $C^3$ function from an open convex domain into a finite-dimensional space ordered by a proper cone $K$. If, for every dual functional $z\in K^*$ and every line $x+th$ staying within the closure of the domain, the scalar function $t\mapsto\langle z,f(x+th)\rangle$ is operator concave on $(-1,1)$, then $f$ is $(K,1)$-compatible with its closed domain. Compatibility is exactly the third-derivative inequality $D^3f(x)[h,h,h]\preceq_K -3D^2f(x)[h,h]$ needed for Nesterov and Nemirovskii's composition lemma to turn a self-concordant barrier of the domain into a self-concordant barrier of the hypograph. Applying this to the sandwiched Rényi quasi-relative entropy $\Psi_\alpha$ yields logarithmic barriers for the hypograph when $\alpha\in[1/2,1]$ and for the epigraph when $\alpha\in[1,2]$, each with the optimal barrier parameter $1+2n$; the perspective of the sandwiched Rényi entropy also obtains a $(2+2n)$-barrier for its epigraph when $\alpha\in[1/2,1)$.

Load-bearing premise

The whole $\alpha\in[1/2,1]$ half of the main theorem rests on a quoted analytic-continuation result, not proved in the paper: a certain trace function formed from the two matrices and their perturbation directions must extend to the upper half-plane in a positivity-preserving way, and if that extension fails on the exact boundary of the stated domain, the hypograph barrier for this range collapses.

Editorial extensions

If this is right

For every $\alpha\in[1/2,1]$, the natural logarithmic barrier for $\operatorname{cl}\operatorname{hypo}\Psi_\alpha$ is a $(1+2n)$-logarithmically homogeneous self-concordant barrier with optimal parameter, so maximizing $\Psi_\alpha$ over convex sets can be handled by interior-point methods.
For every $\alpha\in[1,2]$, the analogous barrier for $\operatorname{cl}\operatorname{epi}\Psi_\alpha$ is also $(1+2n)$-logarithmically homogeneous and optimal, covering the convex minimization range of $\Psi_\alpha$.
For $\alpha\in[1/2,1)$, the perspective of the sandwiched Rényi entropy has a $(2+2n)$-self-concordant barrier for its epigraph, again optimal.
The compatibility results for noncommutative perspectives of operator concave functions previously established in the quantum relative entropy literature follow as corollaries of the one-dimensional criterion.
The barrier functions are implemented in the open-source interior-point solver QICS and are exposed through PICOS; numerical experiments on Rényi mutual information and quantum rate-distortion problems reach residuals between $10^{-8}$ and $10^{-10}$.

Reading between the lines

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

Extending the paper's logic, any future function whose one-dimensional slices can be shown operator concave through Loewner-type arguments would inherit compatibility without a separate third-derivative calculation.
The same strategy could plausibly cover the general trace family $\Psi_{p,q,s}$ in the parameter ranges listed in the conclusion, once an analogue of the quoted continuation lemma is available for those parameters.
If the paper's conjecture for $\alpha\in(2,\infty)$ is true, epigraph barriers with parameter growing like $O(\alpha^3)$ would follow from the same composition lemma, giving an interior-point route for a range that is currently open.

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.

Desk Editor's Note

read the letter

Have you seen this? It actually settles something concrete: no self-concordant barrier was known for the sandwiched Rényi entropy, and here they build optimal-parameter barriers for the hypograph for α∈[1/2,1] and the epigraph for α∈[1,2]. The route is a new sufficient condition that is strictly weaker than the operator-convexity assumptions in earlier compatibility results. That alone makes it worth reading.

The main theorem, 3.1, is cleaner than I expected from the abstract. If every scalar restriction t↦⟨z,f(x+th)⟩ is operator concave on the relevant line segment, then f is (K,1)-compatible. The proof uses the integral representation of operator concave functions and a simple bound on ξ''' in terms of ξ''; I rechecked the constant and it's right. Combining with Nesterov's barrier composition gives the self-concordance claim. The application sections are more involved but the logic holds together. The α∈[1,2] case uses a clever Kronecker-product representation of the trace function and a positive linear map to reduce to concavity of a perspective; I did not find a gap. The α∈[1/2,1] case leans on an analytic-continuation result of Hiai (Lemma 4.2) that is quoted as an intermediate step in the proof of Hiai's Theorem 2.1. That is the softest point. It is load-bearing for the hypograph barrier, and the paper does not reproduce the proof. I could not see a mismatch with the stated matrix ranges or the α interval, but a referee should insist on a precise statement or a short proof in an appendix. This is not a fatal flaw, but it is the one place where I would want more support.

The numerical experiments are a modest but real plus: they solve the Rényi mutual information and a rate-distortion problem, and the residuals match the fixed-point characterization (for one case) and the closed-form limit (for the other). The code is available through QICS, though not pinned to a specific commit; that is a minor reproducibility nit.

The open questions at the end are honest. The conjecture for α∈(2,∞) with compatibility parameter (2α−1)/3 is clearly stated and the scalar case is proven tight. That is a useful pointer for follow-up work.

Who should read this: anyone working on self-concordant barriers, conic optimization with quantum information constraints, or nonsymmetric cones. It deserves a serious referee. I would send it out and expect that a careful referee will ask for a proof or precise statement of Lemma 4.2 and a pinned software version, but nothing more structural.

Referee Report

0 major / 5 minor

Summary. The paper develops a sufficient condition for the natural logarithmic barrier associated with the epigraph (or hypograph) of a function to be self-concordant. The main technical result, Theorem 3.1, shows that if every scalarized restriction of a function along lines is operator concave, then the function is (K,1)-compatible; combining this with Nesterov’s compatibility theorem yields self-concordant barriers. The authors apply this to the sandwiched Rényi quasi-relative entropy Psi_alpha, obtaining a (1+2n)-logarithmically homogeneous self-concordant barrier for the hypograph when alpha in [1/2,1] and for the epigraph when alpha in [1,2], together with optimality of the barrier parameter. They also obtain a barrier for the epigraph of the perspective D_alpha for alpha in [1/2,1), prove a general compatibility result for free mappings (Theorem 4.5), and provide derivative oracles and numerical experiments in the QICS solver.

Significance. The central result is significant: it gives the first self-concordant barriers for the sandwiched Rényi entropy cones, thereby enabling interior-point methods for optimization problems involving these quantum-information-theoretic functions. The line-operator-concavity criterion in Theorem 3.1 is simple, broadly applicable, and unifies previous compatibility results for operator convex functions and their perspectives. The paper is commendable for providing machine-verifiable derivative formulas, an open-source implementation, and numerical verification against known fixed-point and rate-distortion formulas. The main proof is generally clean; the principal external dependency is Lemma 4.2, quoted from Hiai, which I found no evidence to doubt, though it is a verification burden that should be made more explicit.

minor comments (5)

[Section 4.1, Lemma 4.2] Lemma 4.2 is load-bearing for the alpha in [1/2,1] hypograph barrier, but its proof is only a citation to an intermediate step in the proof of Hiai's Theorem 2.1. I did not find a mismatch with the stated ranges or endpoint conditions, but the dependence would be much easier to verify if the authors either reproduced the Pick-function argument or stated the precise theorem in Hiai's notation and explicitly checked the parameter mapping (p=(1-alpha)/(2alpha), q=1, s=alpha) and the semidefinite endpoint constraints X±H >= 0 and Y±V >= 0.
[Appendix A.5] The proof of Proposition 6.1 states that (alpha-2)xhat - (alpha+1)yhat <= 2alpha-1, but the bound needed to derive beta=(2alpha-1)/3 is the lower bound (alpha-2)xhat - (alpha+1)yhat >= -(2alpha-1). The displayed inequality is true but does not by itself imply the required estimate on the third derivative; please correct this step.
[Lemma 2.2 and Theorem 3.1] Lemma 2.2 states that the measure mu is a unique positive finite Borel measure, while the proof of Theorem 3.1 refers to a unique Borel probability measure. These statements are inconsistent when g''(0)=0; please clarify the normalization and the uniqueness claim.
[Lemma 4.4] The statement of Lemma 4.4 ends with 'then f is also H^n_+-convex (a,c)', which should read 'H^n_+-convex on (a,c)'.
[Theorem 1.1 proof] The proof says the result follows directly from Theorem 3.1 and Lemma 2.4. A reader may be confused because Lemma 2.4 contains a factor beta/3; the argument becomes explicit if one applies Lemma 2.4 to the 3-scaled domain barrier 3G, which is a valid (3nu)-self-concordant barrier. Please spell this out.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the self-concordance claims follow from external operator-concavity and analytic-continuation results together with Nesterov's compatibility theorem, not from the paper's own conclusions.

full rationale

The derivation chain is self-contained in the relevant sense. Theorem 3.1 proves that line-wise operator concavity implies (K,1)-compatibility using only the integral representation of operator convex functions and elementary differentiation; it does not presuppose compatibility or self-concordance. The line-wise operator concavity of F is established for alpha in [1/2,1] by Lemma 4.2, quoted as an intermediate result of Hiai's published theorem, and for alpha in [1,2] by the noncommutative perspective construction in Theorem 4.5 and Corollary 4.8, which builds on standard independently established concavity results. Lemma 2.4 is Nesterov's composition theorem, converting compatibility plus domain barriers into self-concordant barriers for epigraphs and hypographs. No parameter is fitted to data and no quantity is renamed as a prediction: the barrier parameters 1+2n and 2+2n are computed explicitly from the -log det barriers and the +1 contribution from the t-direction. The optimality lower bounds cite [4, Corollary 3.13] and [4, Proposition 3.11], a same-author publication, but these are general theorems with independent published proofs applied to the cones in question; under the review rules this is real evidence and does not raise the circularity score. The numerical experiments are consistency checks against external closed-form or fixed-point characterizations. The open questions and unproved remarks, such as Remark 4.11 and Conjecture 6.2, are explicitly labeled and are not load-bearing for Theorems 1.2 and 1.3. The only external dependency worth independent verification is Hiai's analytic-continuation lemma, but citing an external theorem is not a circular step.

Assumptions & free parameters 0 free parameters · 6 assumptions · 0 invented entities

The paper introduces no fitted constants and no new physical or mathematical entities. All quantitative claims are derived from operator-theoretic theorems and prior convexity results; the only input chosen by the authors is the functional form of the barrier, which is the natural log-barrier.

assumptions (6)

standard math Loewner's theorem and the integral representation of operator convex functions (Lemmas 2.1 and 2.2)
Used in the proof of Theorem 3.1 to turn operator concavity along lines into the third-derivative inequality defining compatibility.
standard math Hansen-Tomiyama matrix criterion for Hn+-convex functions (Lemma A.2)
Used in Lemma 4.4 to glue operator concavity on (-1,0) and (0,1) across t=0.
standard math Nesterov-Nemirovskii compatibility composition theorem (Lemma 2.4, from [31, Theorem 5.4.4])
Converts (K,beta)-compatibility plus barriers for the domain and the cone into a self-concordant barrier for the hypograph.
domain assumption Hiai's analytic-continuation theorem for the transpose of Psi_alpha (Lemma 4.2, [12, Theorem 2.1])
Load-bearing for the alpha in [1/2,1] hypograph result; it supplies the Pick-function property that yields operator monotonicity of the transpose and hence operator concavity of F along lines.
domain assumption Known convexity and concavity of Psi_alpha over the full parameter ranges ([9], [5,6]) and of noncommutative perspectives ([34], [35])
Used in Corollary 4.8 and Theorem 4.5 to establish joint Hn+-concavity of the composed perspectives, which then gives compatibility for alpha in [1,2].
domain assumption Lower bound on barrier parameters for homogeneous convex cones ([4, Corollary 3.13])
Used to certify that the (1+2n) and (2+2n) barrier parameters in Theorems 1.2 and 1.3 are optimal.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Operator convexity along lines, self-concordance, and sandwiched R\'enyi entropies." pith.science (2026). https://pith.science/paper/ERAN6SGM

@misc{pith2026250205627,
  author       = {Pith},
  title        = {Pith review of: Operator convexity along lines, self-concordance, and sandwiched R\'enyi entropies},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ERAN6SGM}},
  note         = {Machine review of arXiv:2502.05627}
}

read the original abstract

Barrier methods play a central role in the theory and practice of convex optimization. One of the most general and successful analyses of barrier methods for convex optimization, due to Nesterov and Nemirovskii, relies on the notion of self-concordance. While an extremely powerful concept, proving self-concordance of barrier functions can be very difficult. In this paper we give a simple way to verify that the natural logarithmic barrier of a convex nonlinear constraint is self-concordant via the theory of operator convex functions. Namely, we show that if a convex function is operator convex along any one-dimensional restriction, then the natural logarithmic barrier of its epigraph is self-concordant. We apply this technique to construct self-concordant barriers for the epigraphs of functions arising in quantum information theory. Notably, we apply this to the sandwiched R\'enyi entropy function, for which no self-concordant barrier was known before. Additionally, we utilize our sufficient condition to provide simplified proofs for previously established self-concordance results for the noncommutative perspective of operator convex functions. An implementation of the convex cones considered in this paper is now available in our open source interior-point solver QICS.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

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

A Linearly Convergent Algorithm for Computing the Petz-Augustin Mean
quant-ph 2025-02 conditional novelty 6.0 of 10

A fixed-point iteration computes the Petz-Augustin mean with linear convergence in the Thompson metric for alpha > 1/2, giving the first non-asymptotic guarantees for this quantity and for the Petz capacity.

Reference graph

Works this paper leans on

42 extracted references · 38 canonical work pages · cited by 1 Pith paper

[1]

Nesterov and A

Y. Nesterov and A. Nemirovskii, Interior-point polynomial algorithms in convex program- ming. SIAM, 1994

work page 1994
[2]

Matrix monotonicity and self-c oncordance: how to handle quantum entropy in optimization problems,

L. Faybusovich and T. Tsuchiya, “Matrix monotonicity and self-c oncordance: how to handle quantum entropy in optimization problems,” Optimization Letters , vol. 11, pp. 1513–1526, 2017

work page 2017
[3]

Conic optimization with sp ectral functions on Euclidean Jordan algebras,

C. Coey, L. Kapelevich, and J. P. Vielma, “Conic optimization with sp ectral functions on Euclidean Jordan algebras,” Mathematics of Operations Research , vol. 48, no. 4, pp. 1906–1933, 2023

work page 1906
[4]

Optimal self-concordant barrie rs for quantum relative en- tropies,

H. Fawzi and J. Saunderson, “Optimal self-concordant barrie rs for quantum relative en- tropies,” SIAM Journal on Optimization , vol. 33, no. 4, pp. 2858–2884, 2023

work page 2023
[5]

On quantum R´ enyi entropies: A new generalization and some properties,

M. M¨ uller-Lennert, F. Dupuis, O. Szehr, S. Fehr, and M. Tomam ichel, “On quantum R´ enyi entropies: A new generalization and some properties,” Journal of Mathematical Physics , vol. 54, no. 12, 2013

work page 2013
[6]

Strong converse for the cla ssical capacity of entanglement-breaking and Hadamard channels via a sandwiched R´ enyi relative entropy,

M. M. Wilde, A. Winter, and D. Yang, “Strong converse for the cla ssical capacity of entanglement-breaking and Hadamard channels via a sandwiched R´ enyi relative entropy,” Communications in Mathematical Physics , vol. 331, pp. 593–622, 2014

work page 2014
[7]

Quantum hypothesis testing and the operational interpretation of the quantum R´ enyi relative entropies,

M. Mosonyi and T. Ogawa, “Quantum hypothesis testing and the operational interpretation of the quantum R´ enyi relative entropies,” Communications in Mathematical Physics , vol. 334, pp. 1617–1648, 2015

work page 2015
[8]

Privacy ampliﬁcation and decoupling without smoothing ,

F. Dupuis, “Privacy ampliﬁcation and decoupling without smoothing ,” IEEE Transactions on Information Theory , 2023

work page 2023

Show all 42 references

[9]

Monotonicity of a relative R´ enyi ent ropy,

R. L. Frank and E. H. Lieb, “Monotonicity of a relative R´ enyi ent ropy,” Journal of Mathe- matical Physics, vol. 54, no. 12, 2013

2013
[10]

Remarks on two theorems of E. Lieb,

H. Epstein, “Remarks on two theorems of E. Lieb,” Communications in Mathematical Physics, vol. 31, pp. 317–325, 1973

1973
[11]

Concavity of certain matrix trace functions,

F. Hiai, “Concavity of certain matrix trace functions,” Taiwanese Journal of Mathematics , vol. 5, no. 3, pp. 535–554, 2001

2001
[12]

Concavity of certain matrix trace and norm functions,

——, “Concavity of certain matrix trace and norm functions,” Linear algebra and its ap- plications, vol. 439, no. 5, pp. 1568–1589, 2013

2013
[13]

Concavity of certain matrix trace and norm functions. II ,

——, “Concavity of certain matrix trace and norm functions. II ,” Linear Algebra and its Applications, vol. 496, pp. 193–220, 2016

2016
[14]

From Wigner-Yanase-Dyson conjecture to Carlen -Frank-Lieb conjecture,

H. Zhang, “From Wigner-Yanase-Dyson conjecture to Carlen -Frank-Lieb conjecture,” Ad- vances in Mathematics , vol. 365, p. 107053, 2020

2020
[15]

S. P. Boyd and L. Vandenberghe, Convex optimization . Cambridge university press, 2004

2004
[16]

Minimizing quantum R´ enyi d ivergences via mirror descent with Polyak step size,

J.-K. You, H.-C. Cheng, and Y.-H. Li, “Minimizing quantum R´ enyi d ivergences via mirror descent with Polyak step size,” in 2022 IEEE International Symposium on Information Theory (ISIT) . IEEE, 2022, pp. 252–257

2022
[17]

Simpler semideﬁnite programs for completely boun ded norms,

J. Watrous, “Simpler semideﬁnite programs for completely boun ded norms,” Chicago Jour- nal of Theoretical Computer Science , vol. 2013, no. 8, July 2013

2013
[18]

A matrix convexity approach to some celebrated quantum inequalities,

E. G. Eﬀros, “A matrix convexity approach to some celebrated quantum inequalities,” Proceedings of the National Academy of Sciences , vol. 106, no. 4, pp. 1006–1008, 2009. 19

2009
[19]

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

1986
[20]

Convex trace functions and the Wigner-Yanase-D yson conjecture,

E. H. Lieb, “Convex trace functions and the Wigner-Yanase-D yson conjecture,” Les ren- contres physiciens-math´ ematiciens de Strasbourg-RCP25, vol. 19, pp. 0–35, 1973

1973
[21]

Concavity of certain maps on positive deﬁnite matrice s and applications to Hadamard products,

T. Ando, “Concavity of certain maps on positive deﬁnite matrice s and applications to Hadamard products,” Linear algebra and its applications , vol. 26, pp. 203–241, 1979

1979
[22]

Exploiting structure in qu antum relative entropy programs,

K. He, J. Saunderson, and H. Fawzi, “Exploiting structure in qu antum relative entropy programs,” arXiv preprint arXiv:2407.00241 , 2024

2024 arXiv
[23]

QICS: Quantum information conic solver,

——, “QICS: Quantum information conic solver,” arXiv preprint arXiv:2410.17803 , 2024

2024 arXiv
[24]

Eﬃcient implementation of interior-po int methods for quantum relative entropy,

M. Karimi and L. Tun¸ cel, “Eﬃcient implementation of interior-po int methods for quantum relative entropy,” INFORMS Journal on Computing , 2024

2024
[25]

Performance enhanc ements for a generic conic interior point algorithm,

C. Coey, L. Kapelevich, and J. P. Vielma, “Performance enhanc ements for a generic conic interior point algorithm,” Mathematical Programming Computation , vol. 15, no. 1, pp. 53– 101, 2023

2023
[26]

Semideﬁnite appro ximations of the matrix logarithm,

H. Fawzi, J. Saunderson, and P. A. Parrilo, “Semideﬁnite appro ximations of the matrix logarithm,” Foundations of Computational Mathematics , vol. 19, pp. 259–296, 2019

2019
[27]

Lieb’s concavity theorem, matr ix geometric means, and semideﬁnite optimization,

H. Fawzi and J. Saunderson, “Lieb’s concavity theorem, matr ix geometric means, and semideﬁnite optimization,” Linear Algebra and its Applications , vol. 513, pp. 240–263, 2017

2017
[28]

Inequalities for quan tum divergences and the Audenaert–Datta conjecture,

E. A. Carlen, R. L. Frank, and E. H. Lieb, “Inequalities for quan tum divergences and the Audenaert–Datta conjecture,” Journal of Physics A: Mathematical and Theoretical , vol. 51, no. 48, p. 483001, 2018

2018
[29]

Matrix analysis: matrix monotone functions, matrix mea ns, and majorization,

F. Hiai, “Matrix analysis: matrix monotone functions, matrix mea ns, and majorization,” Interdisciplinary Information Sciences , vol. 16, no. 2, pp. 139–248, 2010

2010
[30]

Simon, Loewner’s theorem on monotone matrix functions

B. Simon, Loewner’s theorem on monotone matrix functions . Springer, 2019, vol. 10

2019
[31]

Nesterov, Lectures on Convex Optimization

Y. Nesterov, Lectures on Convex Optimization . Springer, 2018

2018
[32]

R. G. Bartle, The elements of integration and Lebesgue measure . John Wiley & Sons, 2014

2014
[33]

Proper analytic free m aps,

J. W. Helton, I. Klep, and S. McCullough, “Proper analytic free m aps,” Journal of Func- tional Analysis , vol. 260, no. 5, pp. 1476–1490, 2011

2011
[34]

Perspective s of matrix convex functions,

A. Ebadian, I. Nikoufar, and M. Eshaghi Gordji, “Perspective s of matrix convex functions,” Proceedings of the National Academy of Sciences , vol. 108, no. 18, pp. 7313–7314, 2011

2011
[35]

Diﬀerent quantum f-divergences and t he reversibility of quantum operations,

F. Hiai and M. Mosonyi, “Diﬀerent quantum f-divergences and t he reversibility of quantum operations,” Reviews in Mathematical Physics , vol. 29, no. 07, p. 1750023, 2017

2017
[36]

PICOS: A Python interface to co nic optimization solvers,

G. Sagnol and M. Stahlberg, “PICOS: A Python interface to co nic optimization solvers,” Journal of Open Source Software , vol. 7, no. 70, p. 3915, Feb. 2022

2022
[37]

N. J. Higham, Functions of matrices: theory and computation . SIAM, 2008

2008
[38]

Hiai and D

F. Hiai and D. Petz, Introduction to matrix analysis and applications . Springer Science & Business Media, 2014

2014
[39]

Bhatia, Matrix analysis

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

2013
[40]

Correlation detection and an op erational interpretation of the R´ enyi mutual information,

M. Hayashi and M. Tomamichel, “Correlation detection and an op erational interpretation of the R´ enyi mutual information,” Journal of Mathematical Physics , vol. 57, no. 10, 2016. 20

2016
[41]

Eﬃcient computation of t he quantum rate-distortion function,

K. He, J. Saunderson, and H. Fawzi, “Eﬃcient computation of t he quantum rate-distortion function,” Quantum, vol. 8, p. 1314, Apr. 2024

2024
[42]

Lifting for sim- plicity: Concise descriptions of convex sets,

H. Fawzi, J. Gouveia, P. A. Parrilo, J. Saunderson, and R. R. Th omas, “Lifting for sim- plicity: Concise descriptions of convex sets,” SIAM Review , vol. 64, no. 4, pp. 866–918, 2022. 21

2022

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.

[1] [1]

Nesterov and A

Y. Nesterov and A. Nemirovskii, Interior-point polynomial algorithms in convex program- ming. SIAM, 1994

work page 1994

[2] [2]

Matrix monotonicity and self-c oncordance: how to handle quantum entropy in optimization problems,

L. Faybusovich and T. Tsuchiya, “Matrix monotonicity and self-c oncordance: how to handle quantum entropy in optimization problems,” Optimization Letters , vol. 11, pp. 1513–1526, 2017

work page 2017

[3] [3]

Conic optimization with sp ectral functions on Euclidean Jordan algebras,

C. Coey, L. Kapelevich, and J. P. Vielma, “Conic optimization with sp ectral functions on Euclidean Jordan algebras,” Mathematics of Operations Research , vol. 48, no. 4, pp. 1906–1933, 2023

work page 1906

[4] [4]

Optimal self-concordant barrie rs for quantum relative en- tropies,

H. Fawzi and J. Saunderson, “Optimal self-concordant barrie rs for quantum relative en- tropies,” SIAM Journal on Optimization , vol. 33, no. 4, pp. 2858–2884, 2023

work page 2023

[5] [5]

On quantum R´ enyi entropies: A new generalization and some properties,

M. M¨ uller-Lennert, F. Dupuis, O. Szehr, S. Fehr, and M. Tomam ichel, “On quantum R´ enyi entropies: A new generalization and some properties,” Journal of Mathematical Physics , vol. 54, no. 12, 2013

work page 2013

[6] [6]

Strong converse for the cla ssical capacity of entanglement-breaking and Hadamard channels via a sandwiched R´ enyi relative entropy,

M. M. Wilde, A. Winter, and D. Yang, “Strong converse for the cla ssical capacity of entanglement-breaking and Hadamard channels via a sandwiched R´ enyi relative entropy,” Communications in Mathematical Physics , vol. 331, pp. 593–622, 2014

work page 2014

[7] [7]

Quantum hypothesis testing and the operational interpretation of the quantum R´ enyi relative entropies,

M. Mosonyi and T. Ogawa, “Quantum hypothesis testing and the operational interpretation of the quantum R´ enyi relative entropies,” Communications in Mathematical Physics , vol. 334, pp. 1617–1648, 2015

work page 2015

[8] [8]

Privacy ampliﬁcation and decoupling without smoothing ,

F. Dupuis, “Privacy ampliﬁcation and decoupling without smoothing ,” IEEE Transactions on Information Theory , 2023

work page 2023