Pith. sign in

REVIEW 1 major objections 4 minor 77 references

Information geometry for types in the large-$n$ limit of random matrices

T0 review · 1 major / 4 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read The paper proves that several free entropy quantities—1-bounded entropy, free entropy dimension, and (with endpoint-dependent shifts) microstates free entropy—are bounded below along any Wasserstein geodesic of full types, and gives an…

desk verdict Substantial paper with real new results, but the uniqueness proof for Gibbs types (Prop 5.7) has a genuine gap that needs fixing before the moment-measure claims are solid. read the letter →

arxiv 2501.00703 v2 pith:VULPMPBX submitted 2025-01-01 math.OA math.FAmath.LOmath.PR

David Jekel This is my paper · ORCID
classification math.OAmath.FAmath.LOmath.PR MSC 46L5460B2046L10
keywords freeentropyWassersteingeodesicsfulltypesdefinablepredicatestracialvonNeumannalgebrasrandommatrixmodels1-boundedmomentmeasures
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 aims to bring the geometry of optimal transport to free probability by proving that, for full types—distributions that test all definable formulas, not just polynomials—the free entropy quantities do not decrease along Wasserstein geodesics. Specifically, along the interpolation $x_t=(1-t)x+ty$ of an optimal coupling, the 1-bounded entropy and free entropy dimension of the type of $x_t$ are at least those of either endpoint, and the microstates free entropy is at least the endpoint value minus a dimension-dependent logarithmic penalty. This matters because these quantities describe the large-$n$ limit of invariant random multi-matrix models, so geodesic lower bounds constrain how entropy can behave when matrix models are optimally transported. The paper also establishes existence and uniqueness of Gibbs types and quasi-moment types for strongly convex definable predicates, and gives a counterexample showing that for the usual notion of non-commutative law, entropy and Wasserstein distance cannot be simultaneously realized by random matrix approximations.

What carries the argument

The central objects are full types: linear functionals on all formulas built from trace polynomials by suprema, infima, and continuous connectives, rather than only on non-commutative polynomials as in the usual non-commutative law. The load-bearing machinery is the Monge-Kantorovich duality for full types (Theorem 3.19), which provides convex definable predicates $\varphi,\psi$ whose gradients are definable Lipschitz functions; these gradients act as transport maps between points along a Wasserstein geodesic, and the lifting Lemma 4.9 converts them into measurable selections for random matrix models. The entropy quantities are microstate-based: $\chi$ measures Lebesgue-volume growth of microstate spaces, $\delta$ measures covering-number growth, and $\operatorname{Ent}$ measures orbital covering entropy, all defined through neighborhoods of types.

What would settle it

Compute the full-type free entropy of the geodesic midpoint for the pair of laws constructed in Proposition 4.12, viewed as full types; if the midpoint value falls below $\max(\chi^U_{\mathrm{full}}(\mu)+2m\log(1-t), \chi^U_{\mathrm{full}}(\nu)+2m\log t)$, then Theorem 1.1(3) is false. Alternatively, exhibit an optimal coupling whose Monge-Kantorovich potentials cannot be chosen definable with Lipschitz gradients, which would break Lemma 4.9.

Watch

Extended reading notes

Core claim

For any optimal coupling $(x,y)$ of full types $\mu,\nu\in S_m(T_U)$, the interpolation $\mu_t = \operatorname{tp}((1-t)x+ty)$ satisfies $\operatorname{Ent}^U_{\mathrm{full}}(\mu_t) \geq \max(\operatorname{Ent}^U_{\mathrm{full}}(\mu), \operatorname{Ent}^U_{\mathrm{full}}(\nu))$, $\delta^U_{\mathrm{full}}(\mu_t) \geq \max(\delta^U_{\mathrm{full}}(\mu), \delta^U_{\mathrm{full}}(\nu))$, and $\chi^U_{\mathrm{full}}(\mu_t) \geq \max(\chi^U_{\mathrm{full}}(\mu)+2m\log(1-t), \chi^U_{\mathrm{full}}(\nu)+2m\log t)$. The proof runs through Monge-Kantorovich duality for types: convex definable predicates produce Lipschitz definable transport maps expressing endpoints as functions of interior points, and a lifting lemma extends any random matrix model of an endpoint to one of the geodesic point. The author also shows that Wasserstein-generic full types have $\chi^U_{\mathrm{full}}=-\infty$ and $\operatorname{Ent}^U_{\mathrm{full}}=+\infty$, constructs Gibbs types for strongly convex definable predicates satisfying a Talagrand inequality, and proves existence and uniqueness of quasi-moment types as maximizers of $\chi^U_{\mathrm{full}}(\nu)-C_{\mathrm{full}}(\mu,\nu)-t(\nu,q)$. A separate counterexample shows that in the setting of non-commutative laws (quantifier-free types), no random matrix sequence can simultaneously realize the correct free entropy and the correct Connes-embeddable Wasserstein distance.

Load-bearing premise

The load-bearing premise is that Monge-Kantorovich duality for full types yields convex definable predicates whose gradients are definable Lipschitz functions with controlled operator norm, and that these survive as measurable selections in Lemma 4.9; the author notes in Remark 4.17 that the analogous lifting fails for quantifier-free types.

Editorial extensions

If this is right

  • For any invariant random multi-matrix model approximating the endpoints of a geodesic, the entropy of the interpolated model is bounded below by the endpoint entropies, up to the logarithmic term for $\chi$.
  • Wasserstein-generic full types have $\chi^U_{\mathrm{full}}=-\infty$ and $\operatorname{Ent}^U_{\mathrm{full}}=+\infty$, so finite entropy and infinite 1-bounded entropy occur on dense sets for different reasons.
  • Strongly convex definable potentials admit unique Gibbs types, and these satisfy a Talagrand transportation-cost inequality; consequently, the collection of such Gibbs types is Wasserstein-separable.
  • The quadratically regularized Santambrogio variational problem has a unique solution $\nu_t$ for every type $\mu$, obtained as the Gibbs type of $\varphi_t+tq$ and arising as the Wasserstein limit of the $\varepsilon$-regularized maximizers.
  • For ordinary non-commutative laws, the same simultaneous entropy-distance approximation is impossible in general: Proposition 4.12 constructs laws $\mu,\nu$ with finite free entropy that no uniformly bounded random matrix tuples can realize with both the correct entropy and the correct Connes-embeddable Wasserstein distance.

Reading between the lines

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

  • This suggests that any random matrix approximation scheme for free optimal transport must track logical formulas beyond polynomials; quantified formulas are not a technical nuisance but an essential part of the limiting object.
  • A testable extension is that, for each $t>0$, the map $\mu\mapsto\nu_t$ from Theorem 1.5 is Wasserstein-continuous, which would give a canonical projection from the non-separable type space onto a separable set of Gibbs types.
  • The quasi-moment-type construction likely points toward a free analog of the Gaussian perturbation of moment measures; sending $t\to 0$ in Theorem 1.5 is the natural next step and might recover an unregularized moment type whenever $\chi^U_{\mathrm{full}}(\nu)>-\infty$.
  • Because the Wasserstein topology on full types is non-separable while strongly convex Gibbs types form a separable set, most full types cannot be definable pushforwards of such Gibbs types, so the free moment-measure analogy can cover only a thin class of distributions.
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

1 major / 4 minor

Summary. The paper develops an information geometry for full types in the ultraproduct of matrix algebras, replacing non-commutative laws by model-theoretic types that assign values to all definable predicates. The main results are: lower bounds for the 1-bounded entropy, free entropy dimension, and Voiculescu free entropy along Wasserstein geodesics between full types (Theorem 1.1 and Proposition 1.2); topological genericity results for these quantities (Proposition 1.3); a counterexample showing that for ordinary non-commutative laws no random matrix model can simultaneously realize the Biane–Voiculescu–Wasserstein distance and the free entropy (Proposition 4.12); and existence and uniqueness of Gibbs types for strongly convex definable predicates, a Talagrand inequality for them, and a regularized Santambrogio variational problem whose solution is a unique quasi-moment type (Propositions 5.7, 5.10, and Theorem 1.5).

Significance. If the arguments are completed, the paper gives the first geodesic entropy inequalities in the multivariate non-commutative setting and a rigorous large-n interpretation of moment measures via full types. The counterexample in §4.4 is a valuable, self-contained obstruction showing that quantifier-free laws are insufficient for simultaneous entropy and distance approximation. The paper is careful to import duality results from [52] with stated assumptions, and the main inequalities have explicit constants and falsifiable statements. However, a load-bearing uniqueness proof for Gibbs types in Proposition 5.7 contains a gap that, as written, leaves the moment-measure half of the paper unsupported.

major comments (1)
  1. [§5.1, Proposition 5.7] The uniqueness proof of Proposition 5.7 is invalid as written. After choosing a separating formula η with ε=|(µ,η)−(ν,η)|>0, the neighborhood O is defined by O={ν' : |(ν',ψ)−(ν,ψ)|<ε/2, |(ν',φ)−(ν,φ)|<ε²/64Lc}, where ψ=φ∘f is a Lipschitz cutoff equal to φ on D_R. The separating formula η is never used in the definition of O. Since ψ may not separate µ and ν (for example, with m=1 and φ(X)=tr(X²), the semicircular type µ and a Bernoulli ±1 type ν have (µ,φ)=(ν,φ)=1 yet are different types), the claimed inclusion Γ_R^(n)(O)⊆{|ψ−(µ,ψ)|≥ε/2} is false. Consequently the exponential smallness of µ^(n)(Γ_R^(n)(O)) is not established, and the inequality χ_U(ν)−(ν,φ)<χ_U(µ)−(µ,φ) does not follow. This gap affects the existence and uniqueness of the Gibbs type, which in turn underpins Claim 6.2, the separability result in §5.3, and the quasi-moment type theorem (Theorem 1.5). The argument can likely be repaired by defining O directly from the separating formula η and using the Lipschitz constant of η in the Herbst concentration bound, but this is not what the text does.
minor comments (4)
  1. [§5.1, Proposition 5.7, proof] The concentration estimate for P(|ψ_{M_n}(X^{(n)})−(µ,ψ)|≥ε/2) is stated with exponent e^{-n²ε²/32c}, but Theorem 5.2(4) would give an exponent involving the Lipschitz constant L of ψ, such as e^{-c n²ε²/(8L²)} (up to a factor of 2); the stated bound does not follow from the cited inequality.
  2. [§5.1, Proposition 5.7, proof] The neighborhood O uses the threshold ε²/64Lc, while the later volume estimate uses ε²/32Lc; these factors should be reconciled.
  3. [§5.1, Proposition 5.7, proof] The separating formula η is introduced and used only to define ε; to make the inclusion Γ_R^(n)(O)⊆{|η−(µ,η)|≥ε/2} valid, the neighborhood O should be defined using η rather than ψ.
  4. [§2.1.1, Eq. (1.1)] The statement that dµ^(n)(X) = (1/Z^(n)) e^{-n²V^(n)(X)}dX would benefit from an explicit statement that dX refers to the normalized Lebesgue measure described earlier in the section, to avoid ambiguity with the usual Lebesgue measure.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity found; the main geodesic and entropy arguments rest on published parameter-free theorems and self-contained computations, while the Proposition 5.7 gap is a correctness issue rather than a circularity.

full rationale

The paper's main derivation chain is not circular. Theorem 1.1 and Proposition 1.2 are built from the previously published, parameter-free Monge-Kantorovich duality for full types, quoted as Theorems 3.19 and 3.20 from [52], together with the earlier microstate-entropy framework from [50,51] and self-contained random-matrix lifting and change-of-variables arguments in Lemmas 4.9 and 4.10. Although [52] is by the same author, it is cited as a published theorem with stated assumptions and it does not assume the present geodesic bounds, so under the review rules it counts as independent evidence rather than circularity. The counterexample in Proposition 4.12 and the entropy computations in Claims 4.13 through 4.16 are carried out in the paper and do not feed fitted parameters back into the conclusion. I also considered the self-citation chain around Theorem 3.19 and Claim 6.1; no step reduces a target inequality to a definition or to a fitted parameter. I do note that Proposition 5.7's uniqueness proof appears to contain a genuine gap: the separating formula eta sets epsilon, but the neighborhood O is defined using psi = phi composed with f, so the inclusion Gamma_R^(n)(O) subset of {|psi_(Mn) - (mu,psi)| >= epsilon/2} need not hold when phi does not separate mu from nu. This is a correctness concern for the moment-measure half of the paper, but it is not a circularity: it does not make the conclusion equal to an input by construction. Hence the circularity score is 0.

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

No free parameters are fitted to data. The mathematical objects (full types, definable predicates, quasi-moment types) are defined in the paper or in prior work with rigorous existence proofs. The main external assumptions are prior theorems in free probability, model theory, and classical optimal transport.

assumptions (5)
  • standard math The ultraproduct Q = prod_{n to U} M_n is countably saturated, so every type and every optimal coupling of two types is realized in Q (Remark 3.17).
    Used throughout to define types and to produce optimal couplings for Theorem 1.1.
  • domain assumption Monge-Kantorovich duality for full types: C_full(mu,nu) is the infimum of (mu,phi)+(nu,psi) over convex definable predicates phi,psi with phi+psi >= Re<x,y> (Theorem 3.19, from [52, Theorem 1.1]).
    This is the backbone of Lemma 4.9 and of the quasi-moment type construction in Section 6.
  • domain assumption Gradients of semiconvex and semiconcave definable predicates are definable and L-Lipschitz ([52, Corollary 5.7]), and unitarily equivariant Lipschitz functions satisfy operator norm bounds (Lemma 3.22).
    Permits treating gradients of the interpolated potentials as definable functions on matrix algebras in Sections 3.3 and 4.2.
  • standard math Classical measure-theoretic facts: Rademacher's theorem, change of variables for Lipschitz transformations, and entropy chain rule on finite-dimensional matrix space.
    Used in Lemma 4.10 and Proposition 1.2 to bridge finite-n random matrix entropies to limit type entropy.
  • domain assumption Asymptotic freeness and Connes embeddability of semicircular families, plus Voiculescu's free entropy computations for semicirculars.
    The counterexample in Section 4.4 relies on these standard random matrix and operator algebra facts.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Information geometry for types in the large-$n$ limit of random matrices." pith.science (2026). https://pith.science/paper/VULPMPBX

@misc{pith2026250100703,
  author       = {Pith},
  title        = {Pith review of: Information geometry for types in the large-$n$ limit of random matrices},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VULPMPBX}},
  note         = {Machine review of arXiv:2501.00703}
}
abstract

We study the interaction between entropy and Wasserstein distance in free probability theory. In particular, we give lower bounds for several versions of free entropy dimension along Wasserstein geodesics, as well as study their topological properties with respect to Wasserstein distance. We also study moment measures in the multivariate free setting, showing the existence and uniqueness of solutions for a regularized version of Santambrogio's variational problem. The role of probability distributions in these results is played by types, functionals which assign values not only to polynomial test functions, but to all real-valued logical formulas built from them using suprema and infima. We give an explicit counterexample showing that in the framework of non-commutative laws, the usual notion of probability distributions using only non-commutative polynomial test functions, one cannot obtain the desired large-$n$ limiting behavior for both Wasserstein distance and entropy simultaneously in random multi-matrix models.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

77 extracted references · 54 canonical work pages

  1. [52]

    Optimal transport for types and convex analysis for definable predicates in tracialW∗-algebras

    David Jekel. Optimal transport for types and convex analysis for definable predicates in tracialW∗-algebras. J. Funct. Anal., 287(9):110583, 2024

    work page 2024

  2. [1]

    On non-isomorphic universal sofic groups

    Vadim Alekseev and Andreas Thom. On non-isomorphic universal sofic groups. Preprint, arxiv:2406.06741, 2024

    arXiv 2024

  3. [2]

    An introduction toII1 factors

    Claire Anantharaman-Delaroche and Sorin Popa. An introduction toII1 factors. 2021

    work page 2021

  4. [3]

    Anderson, Alice Guionnet, and Ofer Zeitouni.An Introduction to Random Matrices

    Greg W. Anderson, Alice Guionnet, and Ofer Zeitouni.An Introduction to Random Matrices. Cambridge Studies in Advanced Mathematics. Cambridge University Press, 2009

    2009

  5. [4]

    Vladimir I. Arnold. Sur la g´ eom´ etrie diff´ erentielle des groupes de lie de dimension infinie et ses applications ` a l’hydrodynamique des fluides parfaits.Ann. Inst. Fourier (Grenoble), 16:319–361, 1966. 9The φt here is not directly related to theφε,t above, but obtained from a separate application of the MK duality. INFORMATION GEOMETRY FOR TYPES IN...

    work page 1966

  6. [5]

    Factorial relative commutants and the generalized jung property for ii1 factors.Advances in Mathematics, 396:108107, 2022

    Scott Atkinson, Isaac Goldbring, and Srivatsav Kunnawalkam Elayavalli. Factorial relative commutants and the generalized jung property for ii1 factors.Advances in Mathematics, 396:108107, 2022

    work page 2022

  7. [6]

    On ultraproduct embeddings and amenability for tracial von neumann algebras.International Mathematics Research Notices, 2021(4):2882–2918, 10 2020

    Scott Atkinson and Srivatsav Kunnawalkam Elayavalli. On ultraproduct embeddings and amenability for tracial von neumann algebras.International Mathematics Research Notices, 2021(4):2882–2918, 10 2020

    work page 2021

  8. [7]

    Free moment measures and laws.Journal of Functional Analysis, 284(2):109756, 2023

    Juniper Bahr and Nick Boschert. Free moment measures and laws.Journal of Functional Analysis, 284(2):109756, 2023

    work page 2023

Show all 77 references
  1. [8]

    Bakry and Michel´Emery

    D. Bakry and Michel´Emery. Diffusions hypercontractives. InS´ eminaire de Probabilit´ es XIX, 1983/84, volume 1123 ofLecture Notes in Mathematics, pages 177–206. Springer, Berlin, 1985

    1983

  2. [9]

    Bekerman, Alessio Figalli, and Alice Guionnet

    F. Bekerman, Alessio Figalli, and Alice Guionnet. Transport maps forβ-matrix models and universality.Comm. Math. Phys., 338(2):589–619, 2015

    2015

  3. [10]

    Ben Yaacov, T

    I. Ben Yaacov, T. Ibarluc´ ıa, and T. Tsankov. Extremal models and direct integrals in affine logic. Preprint arXiv:2407.13344, 2024

    2024 arXiv

  4. [11]

    Topometric spaces and perturbations of metric structures.Log Anal, 1:235–272, 2008

    Ita¨i Ben Yaacov. Topometric spaces and perturbations of metric structures.Log Anal, 1:235–272, 2008

    2008

  5. [12]

    On theories of random variables.Israel J

    Ita¨i Ben Yaacov. On theories of random variables.Israel J. Math., 194(2):957–1012, 2013

    2013

  6. [13]

    Ward Henson, and Alexander Usvyatsov

    Ita¨i Ben Yaacov, Alexander Berenstein, C. Ward Henson, and Alexander Usvyatsov. Model theory for metric structures. In Z. Chatzidakis et al., editor,Model Theory with Applications to Algebra and Analysis, Vol. II, volume 350 ofLondon Mathematical Society Lecture Notes Series,...

    2008

  7. [14]

    Continuous first order logic and local stability.Transactions of the American Mathematical Society, 362(10):5213–5259, 10 2010

    Ita¨i Ben Yaacov and Alexander Usvyatsov. Continuous first order logic and local stability.Transactions of the American Mathematical Society, 362(10):5213–5259, 10 2010

    2010

  8. [15]

    Biane, M

    P. Biane, M. Capitaine, and A. Guionnet. Large deviation bounds for matrix Brownian motion.Inventiones Mathematicae, 152:433–459, 2003

    2003

  9. [16]

    Logarithmic sobolev inequalities, matrix models and free entropy.Acta Mathematica Sinica, English Series, 19(3):497–506, 2003

    Philippe Biane. Logarithmic sobolev inequalities, matrix models and free entropy.Acta Mathematica Sinica, English Series, 19(3):497–506, 2003

    2003

  10. [17]

    Concavification of free entropy.Adv

    Philippe Biane and Yoann Dabrowski. Concavification of free entropy.Adv. Math., 234:667–696, 2013

    2013

  11. [18]

    A free probability analogue of the Wasserstein metric on the trace- state space.Geometric and Functional Analysis, 11:1125–1138, 2001

    Philippe Biane and Dan-Virgil Voiculescu. A free probability analogue of the Wasserstein metric on the trace- state space.Geometric and Functional Analysis, 11:1125–1138, 2001

    2001

  12. [19]

    Springer-Verlag, Berlin, Heidelberg, 2006

    Bruce Blackadar.Operator Algebras: Theory ofC ∗-algebras and von Neumann algebras, volume 122 ofEncy- clopaedia of Mathematical Sciences. Springer-Verlag, Berlin, Heidelberg, 2006

    2006

  13. [20]

    S. G. Bobkov and M. Ledoux. From Brunn-Minkowski to Brascamp-Lieb and to logarithmic Sobolev inequalities. Geometric and Functional Analysis, 10:1028–1052, 2000

    2000

  14. [21]

    Brown and Narutaka Ozawa

    Nathaniel P. Brown and Narutaka Ozawa. C∗-algebras and finite-dimensional approximations, volume 88 of Graduate Studies in Mathematics. American Mathematical Society, Providence, 2008

    2008

  15. [22]

    A survey on Connes’ embedding conjecture

    Valerio Capraro. A survey on Connes’ embedding conjecture. arXiv:1003.2076, 2010

    2010 arXiv

  16. [23]

    A quantitative fourth moment theorem in free probability theory.Adv

    Guillaume C´ ebron. A quantitative fourth moment theorem in free probability theory.Adv. Math., 380:107579, 2021

    2021

  17. [24]

    Moment measures.J

    Dario Cordero-Erausquin and Bo’az Klartag. Moment measures.J. Funct. Anal., 268(12):3834–3866, 2015

    2015

  18. [25]

    A non-commutative path space approach to stationary free stochastic differential equations

    Yoann Dabrowski. A non-commutative path space approach to stationary free stochastic differential equations. arxiv:1006.4351, 2010

    2010 arXiv

  19. [26]

    Free transport for convex potentials.New Zealand Journal of Mathematics, 52:259–359, 2021

    Yoann Dabrowski, Alice Guionnet, and Dimitri Shlyakhtenko. Free transport for convex potentials.New Zealand Journal of Mathematics, 52:259–359, 2021

    2021

  20. [27]

    Free Stein kernels and moments maps

    Charles-Philippe Diez. Free Stein kernels and moments maps. Preprint, arXiv:2410.02470, 2024

    2024 arXiv

  21. [28]

    A sharp symmetrized free transport-entropy inequality for the semicircular law

    Charles-Philippe Diez. A sharp symmetrized free transport-entropy inequality for the semicircular law. Preprint, arXiv:2410.02715, 2024

    2024 arXiv

  22. [29]

    Gauthier-Villars, Paris, 2 edition, 1969

    Jacques Dixmier.Les C ∗-alg` ebres et leurs repr´ esentations, volume 29 ofCahiers Scientifiques. Gauthier-Villars, Paris, 2 edition, 1969. Reprinted by Editions Jacques Gabay, Paris, 1996. Translated asC ∗-algebras, North- Holland, Amsterdam, 1977. First Edition 1964

    1969

  23. [30]

    Quantifier elimination inII1 factors

    Ilijas Farah. Quantifier elimination inII1 factors. To appear in M¨ unster J. Math., 2024

    2024

  24. [31]

    Model theory of operator algebras II: model theory.Israel Journal of Mathematics, 201(1):477–505, 2014

    Ilijas Farah, Bradd Hart, and David Sherman. Model theory of operator algebras II: model theory.Israel Journal of Mathematics, 201(1):477–505, 2014

    2014

  25. [32]

    A sharp symmetrized form of talagrand’s transport-entropy inequality for the gaussian measure

    Max Fathi. A sharp symmetrized form of talagrand’s transport-entropy inequality for the gaussian measure. Electron. Commun. Probab., 23(81):1–9, 2018

    2018

  26. [33]

    Stein kernels and moments maps.The Annals of Probability, 47(4):2172–2185, 2019

    Max Fathi. Stein kernels and moments maps.The Annals of Probability, 47(4):2172–2185, 2019

    2019

  27. [34]

    Universality in several-matrix models via approximate transport maps.Acta Math., 217(1):81–176, 2016

    Alessio Figalli and Alice Guionnet. Universality in several-matrix models via approximate transport maps.Acta Math., 217(1):81–176, 2016

    2016

  28. [35]

    Duality for optimal couplings in free probability.Communications in Mathematical Physics, 396:903–981, 2022

    Wilfrid Gangbo, David Jekel, Kyeongsik Nam, and Dimitri Shlyakhtenko. Duality for optimal couplings in free probability.Communications in Mathematical Physics, 396:903–981, 2022

    2022

  29. [36]

    A survey on the model theory of tracial von Neumann algebras

    Isaac Goldbring and Bradd Hart. A survey on the model theory of tracial von Neumann algebras. In Isaac Goldbring, editor,Model Theory of Operator Algebras, pages 133–157. DeGruyter, Berlin, Boston, 2023. 56 DA VID JEKEL

    2023

  30. [37]

    Logarithmic sobolev inequalities.American Journal of Mathematics, 97(4):1061–1083, 1975

    Leonard Gross. Logarithmic sobolev inequalities.American Journal of Mathematics, 97(4):1061–1083, 1975

    1975

  31. [38]

    Combinatorial aspects of random matrix models.Latin American Journal of Probability and Statistics (ALEA), 1:241–279, 2006

    Alice Guionnet and Edouard Maurel-Segala. Combinatorial aspects of random matrix models.Latin American Journal of Probability and Statistics (ALEA), 1:241–279, 2006

    2006

  32. [39]

    On classical analogues of free entropy dimension.J

    Alice Guionnet and Dimitri Shlyakhtenko. On classical analogues of free entropy dimension.J. Func. Anal., 251(2):738–771, 2007

    2007

  33. [40]

    Free diffusions and matrix models with strictly convex interaction

    Alice Guionnet and Dimitri Shlyakhtenko. Free diffusions and matrix models with strictly convex interaction. Geometric and Functional Analysis, 18(6):1875–1916, 03 2009

    1916

  34. [41]

    Free monotone transport.Inventiones Mathematicae, 197(3):613–661, 09 2014

    Alice Guionnet and Dimitri Shlyakhtenko. Free monotone transport.Inventiones Mathematicae, 197(3):613–661, 09 2014

    2014

  35. [42]

    An introduction to continuous model theory

    Bradd Hart. An introduction to continuous model theory. In Isaac Goldbring, editor,Model Theory of Operator Algebras, pages 83–131. DeGruyter, Berlin, Boston, 2023

    2023

  36. [43]

    1-bounded entropy and regularity problems in von Neumann algebras.International Mathematics Research Notices, 1(3):57–137, 2018

    Ben Hayes. 1-bounded entropy and regularity problems in von Neumann algebras.International Mathematics Research Notices, 1(3):57–137, 2018

    2018

  37. [44]

    A random matrix approach to absorption theorems for free products.International Mathematics Research Notices, 2021(3):1919–1979, 2021

    Ben Hayes, David Jekel, Brent Nelson, and Thomas Sinclair. A random matrix approach to absorption theorems for free products.International Mathematics Research Notices, 2021(3):1919–1979, 2021

    2021

  38. [45]

    Free analog of pressure and its legendre transform.Commun

    Fumio Hiai. Free analog of pressure and its legendre transform.Commun. Math. Phys., 255:229–252, 2005

    2005

  39. [46]

    Free transportation cost inequalities for noncommutative multi-variables

    Fumio Hiai and Yoshimichi Ueda. Free transportation cost inequalities for noncommutative multi-variables. Infinite Dimensional Analysis, Quantum Probability, and Related Topics, 9:391–412, 2006

    2006

  40. [47]

    An introduction to von neumann algebras

    Adrian Ioana. An introduction to von neumann algebras. In Isaac Goldbring, editor,Model Theory of Operator Algebras, pages 43–81. DeGruyter, Berlin, Boston, 2023

    2023

  41. [48]

    An elementary approach to free entropy theory for convex potentials.Anal

    David Jekel. An elementary approach to free entropy theory for convex potentials.Anal. PDE, 13(8):2289–2374, 2020

    2020

  42. [49]

    Conditional expectation, entropy, and transport for convex Gibbs laws in free probability.Int

    David Jekel. Conditional expectation, entropy, and transport for convex Gibbs laws in free probability.Int. Math. Res. Not. IMRN, (6):4514–4619, 2022

    2022

  43. [50]

    Covering entropy for types in tracialW∗-algebras

    David Jekel. Covering entropy for types in tracialW∗-algebras. Journal of Logic and Analysis, 15(2):1–68, 2023

    2023

  44. [51]

    Free probability and model theory of tracialW∗-algebras

    David Jekel. Free probability and model theory of tracialW∗-algebras. In Isaac Goldbring, editor,Model Theory of Operator Algebras, pages 215–267. DeGruyter, Berlin, Boston, 2023

    2023

  45. [53]

    Tracial non-commutative smooth functions and the free Wasserstein manifold.Dissertationes Mathematicae, 580:1–150, 2022

    David Jekel, Wuchen Li, and Dimitri Shlyakhtenko. Tracial non-commutative smooth functions and the free Wasserstein manifold.Dissertationes Mathematicae, 580:1–150, 2022

    2022

  46. [54]

    An elementary proof of the inequalityχ ≤ χ∗ for conditional free entropy.Doc

    David Jekel and Jennifer Pi. An elementary proof of the inequalityχ ≤ χ∗ for conditional free entropy.Doc. Math., 29(5):1085–1124, 2024

    2024

  47. [55]

    Zhengfeng Ji, Anand Natarajan, Thomas Vidick, John Wright, and Henry Yuen. MIP*=RE. arXiv:2001.04383, 2020

    2001 arXiv

  48. [56]

    The variational formulation of the Fokker–Planck equation

    Richard Jordan, David Kinderlehrer, and Felix Otto. The variational formulation of the Fokker–Planck equation. SIAM Journal on Mathematical Analysis, 29(1):1–17, 1998

    1998

  49. [57]

    A free entropy dimension lemma.Pacific J

    Kenley Jung. A free entropy dimension lemma.Pacific J. Math., 211(2):265–271, 2003

    2003

  50. [58]

    Strongly1-bounded von Neumann algebras.Geom

    Kenley Jung. Strongly1-bounded von Neumann algebras.Geom. Funct. Anal., 17(4):1180–1200, 2007

    2007

  51. [59]

    Kadison and John R

    Richard V. Kadison and John R. Ringrose.Fundamentals of the Theory of Operator Algebras I, volume 15 of Graduate Studies in Mathematics. American Mathematical Society, Providence, 1997

    1997

  52. [60]

    A general theorem on selectors.Bull

    Kazimierz Kuratowski and Czes law Ryll-Nardzewski. A general theorem on selectors.Bull. Acad. Polon. Sci. S´ er. Sci. Math. Astronom. Phys., 13:397–403, 1965

    1965

  53. [61]

    Lafferty

    John D. Lafferty. The density manifold and configuration space quantization.Transactions of the American Mathematical Society, 305(2):699–741, 1988

    1988

  54. [62]

    Robert J. McCann. A convexity principle for interacting gases.Advances in Mathematics, 128(1):153–179, 1997

    1997

  55. [63]

    Free monotone transport without a trace.Communications in Mathematical Physics, 334(3):1245– 1298, 2015

    Brent Nelson. Free monotone transport without a trace.Communications in Mathematical Physics, 334(3):1245– 1298, 2015

    2015

  56. [64]

    The geometry of dissipative evolution equations the porous medium equation.Communications in Partial Differential Equations, 26(1-2):101–174, 2001

    Felix Otto. The geometry of dissipative evolution equations the porous medium equation.Communications in Partial Differential Equations, 26(1-2):101–174, 2001

    2001

  57. [65]

    Generalization of an inequality by Talagrand and links with the logarithmic Sobolev inequality.Journal of Functional Analysis, 173(2):361–400, 2000

    Felix Otto and C´ edric Villani. Generalization of an inequality by Talagrand and links with the logarithmic Sobolev inequality.Journal of Functional Analysis, 173(2):361–400, 2000

    2000

  58. [66]

    Springer-Verlag, Berlin Heidelberg, 1971

    Shˆ oichirˆ o Sakai.C∗-algebras and W∗-algebras, volume 60 ofErgebnisse der Mathematik und ihrer Grenzgebiete. Springer-Verlag, Berlin Heidelberg, 1971

    1971

  59. [67]

    Dealing with moment measures via entropy and optimal transport.Journal of Functional Analysis, 271(2):418–436, 2016

    Filippo Santambrogio. Dealing with moment measures via entropy and optimal transport.Journal of Functional Analysis, 271(2):418–436, 2016

    2016

  60. [68]

    Fractional free convolution powers.Indiana University Mathematics Journal, To appear

    Dimitri Shlyakhtenko and Terence Tao. Fractional free convolution powers.Indiana University Mathematics Journal, To appear. With an appendix by David Jekel

  61. [69]

    PhD thesis, University of Illinois at Urbana-Champaign, 2011

    Shichang Song.Model theory and probability. PhD thesis, University of Illinois at Urbana-Champaign, 2011. INFORMATION GEOMETRY FOR TYPES IN THE LARGE- n LIMIT OF RANDOM MATRICES 57

    2011

  62. [70]

    Theory of Operator Algebras I, volume 124 of Encyclopaedia of Mathematical Sciences

    Masamichi Takesaki. Theory of Operator Algebras I, volume 124 of Encyclopaedia of Mathematical Sciences. Springer-Verlag, Berlin Heidelberg, 2002

    2002

  63. [71]

    Transportation cost for gaussian and other product measures.Geom

    Michel Talagrand. Transportation cost for gaussian and other product measures.Geom. Funct. Anal., 6:587–600, 1996

    1996

  64. [72]

    Amer- ican Mathematical Society, 2012

    Terence Tao.An Introduction to Random Matrix Theory, volume 132 ofGraduate Texts in Mathematics. Amer- ican Mathematical Society, 2012

    2012

  65. [73]

    Springer, Berlin, 2009

    C´ edric Villani.Optimal Transport: Old and New , volume 338 of Grundlehren Der Mathematischen Wis- senschaften. Springer, Berlin, 2009

    2009

  66. [74]

    The analogues of entropy and of Fisher’s information in free probability, II.Inventiones Mathematicae, 118:411–440, 1994

    Dan-Virgil Voiculescu. The analogues of entropy and of Fisher’s information in free probability, II.Inventiones Mathematicae, 118:411–440, 1994

    1994

  67. [75]

    The analogues of entropy and of Fisher’s information measure in free probability, III: Absence of Cartan subalgebras.Geometric and Functional Analysis, 6:172–199, 1996

    Dan-Virgil Voiculescu. The analogues of entropy and of Fisher’s information measure in free probability, III: Absence of Cartan subalgebras.Geometric and Functional Analysis, 6:172–199, 1996

    1996

  68. [76]

    The analogues of entropy and of Fisher’s information in free probability V.Inventiones Mathematicae, 132:189–227, 1998

    Dan-Virgil Voiculescu. The analogues of entropy and of Fisher’s information in free probability V.Inventiones Mathematicae, 132:189–227, 1998

    1998

  69. [77]

    An Introduction to Operator Algebras

    Kehe Zhu. An Introduction to Operator Algebras. Studies in Advanced Mathematics. CRC Press, Ann Arbor, 1993. Department of Mathematical Sciences, University of Copenhagen, Universitetsparken 5, 2100 Copenhagen Ø, Denmark Email address: daj@math.ku.dk URL: https://davidjekel.com

    1993

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.