Pith. sign in

REVIEW 2 major objections 3 minor 2 cited by

W*-correlations of II$_1$ factors and rigidity of tensor products and graph products

T0 review · 2 major / 3 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read For every subset $F\subset\{2,3,\dots\}$, the group $G_F=\ast_{n\in F}(F_2)^{\times n}$ produces a II$_1$ factor $L(G_F)$ that is W*-correlated only to itself among this family, yielding the first uncountable family of groups that are not…

desk verdict Strong new rigidity results with honest, detailed proofs; one unresolved black-box invocation in Theorem 6.3 that a referee should push on, but the main theorems look solid. read the letter →

arxiv 2507.04691 v2 pith:5TBHA3IS submitted 2025-07-07 math.OA math.GR

Daniel Drimbe , Stefaan Vaes This is my paper · ORCID
classification math.OAmath.GR MSC 46L1046L3646L5446L55
keywords W*-correlationvonNeumannequivalencemeasureII1factorstensorproductrigiditygraphproductsstablysolidgroupalgebras
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 introduces the terminology "W*-correlated" for II$_1$ factors and proves that this very coarse equivalence can separate groups rigidly. For each subset $F\subset\{2,3,\dots\}$, let $G_F$ be the free product of the direct products $(F_2)^{\times n}$ over $n\in F$; the main theorem says that if $F\ne F'$, then $L(G_F)$ and $L(G_{F'})$ are not W*-correlated. Because W*-correlation coincides with von Neumann equivalence for II$_1$ factors, this is the first uncountable family of discrete groups that are pairwise not von Neumann equivalent, and hence also pairwise not measure equivalent and not W*-equivalent. The proof builds two structural rigidity theorems: W*-correlations of tensor products of nonamenable stably solid factors split into factor-by-factor correlations, and W*-correlations of graph products over rigid graphs force matching correlations of link subalgebras.

What carries the argument

The named object carrying the argument is a W*-correlation: a Hilbert $A$-$P\otimes B$-bimodule that is of finite type as a right $P\otimes B$-module, coarse as an $A$-$P$-bimodule, and faithful, with a matching finite-type condition on the left; for factors this is exactly von Neumann equivalence. The proof machinery has three pillars. Stable solidity says that a diffuse subalgebra that does not intertwine into an auxiliary algebra has relatively amenable relative commutant, and the paper proves that $q$-Gaussian factors, certain crossed products, and $(AO)^+$ factors with W$^*$CBAP are stably solid. Rigid graphs satisfy $\operatorname{link}(\operatorname{link} s)=\{s\}$, which forces the normalizer of each link subalgebra to be its star, making link subalgebras behave like tensor-product atoms. Finally, an external dichotomy for subalgebras of amalgamated free products, applied inside $P\otimes B$ decomposed at a vertex, is what converts a W*-correlation of graph products into a W*-correlation of link subalgebras.

What would settle it

The central claim would be falsified by exhibiting distinct subsets $F,F'$ and a faithful W*-correlation between $L(G_F)$ and $L(G_{F'})$. The proof's local bottleneck is Theorem 5.3, so the smallest concrete falsifier would be a W*-correlation between two different tensor powers $L(F_2)^{\otimes n}$ and $L(F_2)^{\otimes m}$ with $n\ne m$; no such correlation is believed to exist, and Theorem 5.3 asserts none does. One could also check directly whether the amalgamated free product dichotomy invoked at the end of Theorem 6.3 satisfies all of its hypotheses for arbitrary rigid graphs.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that W*-correlation is a robust invariant at the level of group von Neumann algebras. A W*-correlation between finite von Neumann algebras $A$ and $B$ is a Hilbert $A$-$P\otimes B$-bimodule of finite type on both sides and coarse on the $A$-$P$ side, for some finite von Neumann algebra $P$; by Proposition 4.8 this is the same as von Neumann equivalence for II$_1$ factors. The main result, Theorem 6.4, is that the factors $L(G_F)$, $G_F=\ast_{n\in F}(F_2)^{\times n}$, are pairwise non-W*-correlated as $F$ runs over all subsets of $\{2,3,\dots\}$. Since measure equivalence and W*-equivalence both imply von Neumann equivalence, these groups are pairwise neither measure equivalent nor W*-equivalent, and they are not virtually isomorphic. The route passes through two structural theorems: Theorem 5.1 says a W*-correlation between a tensor product of nonamenable stably solid factors and an arbitrary tensor product forces a matching partition into W*-correlated sub-tensor-products, and Theorem 6.3 says a W*-correlation between graph products over rigid graphs forces, for each vertex, a W*-correlation between the corresponding link subalgebras.

Load-bearing premise

The load-bearing premise is that an existing dichotomy theorem about von Neumann subalgebras of amalgamated free products applies unchanged to the specific rigid graph products and the specific projections constructed in the proof of Theorem 6.3; the paper invokes that theorem without independently verifying all of its hypotheses in this setting. If that dichotomy fails, the passage from W*-correlated graph products to W*-correlated link subalgebras, and with it the uncountable family construction, collapses.

Editorial extensions

If this is right

  • The groups $G_F$ and $G_{F'}$ with $F\ne F'$ are not von Neumann equivalent, not measure equivalent, and not W*-equivalent, so none of the classical rigidity invariants can put two different members of this family together.
  • For tensor products of nonamenable stably solid factors, W*-correlation is as rigid as isomorphism up to permutation: the number of tensor factors is preserved and matched factors are W*-correlated, giving a W*-correlation version of unique prime factorization.
  • For stably isomorphic graph products over rigid graphs labeled by nonamenable stably solid factors, the underlying graphs must be isomorphic and matching vertices must carry stably isomorphic factors (Theorem 6.5).
  • For merely W*-correlated rigid graph products, the conclusion is weaker: link subalgebras must match pairwise, but the full graphs need not be isomorphic, even for virtually isomorphic examples (Corollary 6.7).

Reading between the lines

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

  • The same free-product construction should produce other uncountable families of pairwise non-von-Neumann-equivalent groups whenever the vertex labels are taken from any nonamenable stably solid factors whose tensor powers are pairwise non-W*-correlated, such as the $q$-Gaussian factors or certain quantum-group factors.
  • Since all diffuse amenable finite von Neumann algebras form a single W*-correlation class, the rigidity displayed here is intrinsically nonamenable; any invariant that separates these groups at the W*-correlation level must be sensitive to the nonamenable, stably solid tensor-factor structure.
  • A plausible testable extension is to replace free group factors by free quantum group factors or right-angled Coxeter group algebras at the vertices, which would yield rigidity for graph products over a larger class of rigid graphs and possibly uncountably many W*-correlation classes coming from the graph itself rather than the labels.
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

2 major / 3 minor

Summary. The paper proposes the terminology "W*-correlation" for the notion previously called measure equivalence of finite von Neumann algebras in [BV22], and studies its relation to von Neumann equivalence from [IPR19]. It proves that for II_1 factors the two notions coincide (Proposition 4.8), and then establishes rigidity results for W*-correlated tensor products (Theorems 5.1 and 5.3) and for graph products over rigid graphs (Theorem 6.3). The main application is Theorem 6.4 / Theorem A: for every subset F of {2,3,...}, with G_F the free product over n in F of the n-fold direct product of F_2, the algebras L(G_F) are pairwise not W*-correlated, hence the groups are pairwise not von Neumann equivalent, not measure equivalent, and not W*-equivalent. The paper also proves a stable-isomorphism rigidity theorem for graph products (Theorem C / 6.5) and a classification of amenable finite von Neumann algebras into three W*-correlation classes (Proposition 4.11).

Significance. The results are substantial. If the main theorem stands, it gives the first uncountable family of pairwise non-von-Neumann-equivalent discrete groups and significantly extends the tensor-product rigidity of [OP03] to the much coarser W*-correlation relation. The paper is rich in technical content: several preparatory results, including Lemma 2.1, Proposition 2.5, and the stable solidity results for q-Gaussian algebras and related crossed products (Theorem 3.2), are proved in detail, and Proposition 4.11 gives a clean classification of amenable W*-correlation classes. The overall architecture of the argument is natural and uses appropriate external tools. However, the central graph-product rigidity theorem depends on a step in the proof of Theorem 6.3 that is not justified as written, and this step is load-bearing for Theorem A.

major comments (2)
  1. [Theorem 6.3, proof, after the definition of p1] The text applies [Ioa12, Theorem 6.4] to the decomposition P⊗B = (P⊗B_star t) ∗_{P⊗B_link t} (P⊗B_{T\{t}}) without stating or verifying the hypotheses of that theorem. The only property established for the corner α(A_s)p1 is strong nonamenability relative to P⊗B_{T\{t}}, whereas the amalgam in the cited theorem is P⊗B_link t. It is not explained why the relative nonamenability required by [Ioa12, Theorem 6.4] holds with respect to P⊗B_link t, nor are possible factoriality, separability, or other hypotheses checked for the relevant corner and the ambient algebras. Since the dichotomy "either α(A_link s)p1 ≺ P⊗B_link t or α(A_star s)p1 ≺ P⊗B_star t" is the decisive step of the proof, this omission is not a mere citation issue.
  2. [Theorem 6.3, proof, first case after the dichotomy] Even granting the dichotomy, the sentence "In the first case, we have found the required t" is not justified. From α(A_link s)p1 ≺ P⊗B_link t for a nonzero p1 ≤ p0 one cannot conclude α(A_link s)p0 ≺ P⊗B_link t; the complementary corner p0−p1 is not treated. The membership p1 ∈ α(A_star s)' ∩ p(P⊗B)p does not make this upgrade automatic. Since Lemma 4.15(i) requires the statement for every p0 ∈ α(A_star s)' ∩ p(P⊗B)p, the proof does not establish the hypotheses of Lemma 4.15. Consequently Theorem 6.3, and with it Theorem 6.4 and Theorem A, is incomplete as written; Theorem 6.5 inherits the same dependence.
minor comments (3)
  1. [§4.4, proof of Proposition 4.11] The sentence "Define c ∈ Z(B) as the support of the support of the right action of B on z·H" appears to contain a duplicated phrase; it should presumably read "support of the right action of B on z·H".
  2. [§4.5, proof of Theorem 4.13(i)] In the display beginning "lim_n (lim sup_k ∥bζ_{n,k} − ζ_{n,k}b∥2)", the right-hand side is missing the "= 0" that the sentence requires.
  3. [§6, Proposition 6.6] In the construction of the graph Γ', the list of vertex labelings is somewhat compressed; it would be clearer to state explicitly that for every i ∈ [k] and s ∈ S \ star s1 the vertex (i,s) is labeled by G_s and is connected according to the stated edge rules.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the rigidity theorems are proved from external dichotomy results and prior independent lemmas; self-citations are not load-bearing and no prediction reduces to a fitted input.

full rationale

The paper's central claims are genuine theorems rather than repackaged inputs. Theorem A follows from Theorem 6.3 and Theorem 5.3, both of which are proved in the text using external results such as [OP03], [Ioa12, Theorem 6.4], and [DV24], together with the paper's own technical lemmas. The notion W*-correlation is a renaming of the 'measure equivalence' bimodule notion from [BV22], but the main rigidity statements do not follow from the renaming; they are derived from solidity, stable solidity, and amalgamated free product dichotomy arguments. There is no fitted parameter that is later called a prediction, no definition that builds in the target conclusion, and no self-citation chain that forces the result. The proof of Theorem 6.3 invokes [Ioa12, Theorem 6.4] as a black box and upgrades an intertwining statement from a corner projection p1 to p0; the text does not fully verify the hypotheses of that dichotomy, and the corner upgrade is not automatic in general. This is a correctness or completeness concern about the proof as written, not a circularity: the claimed conclusion is not equivalent to the cited input by construction. Overall, no significant circularity is present.

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

The paper introduces no free parameters or invented entities. The central claim rests on a network of established theorems in II_1 factor theory (solidity, intertwining-by-bimodules, relative amenability, amalgamated free product rigidity) that are cited as black boxes. The key new property 'stably solid' is defined and proven for the required classes, so it does not enter as an unproved axiom.

assumptions (5)
  • domain assumption Ozawa's solidity theorem and the unique prime factorization theorem for tensor products of solid factors [Oza03, OP03] are valid and apply to the classes used.
    Used in Section 5 to establish the framework for tensor product rigidity; these are prior published theorems.
  • domain assumption The theorem of Ioana on commuting subalgebras in amalgamated free products [Ioa12, Theorem 6.4] applies with the stated hypotheses in the proof of Theorem 6.3.
    The dichotomy in the proof of Theorem 6.3 depends on this external result; the hypotheses are not re-verified in the paper.
  • domain assumption The deformation estimates for q-Gaussian algebras stated in Proposition 2.3 are complete and correct, relying on [CIW19, Proposition 4.9] and [Wil20, Proposition 6.6].
    The paper cites that the original proof in [Avs11] was incomplete and uses these external results; this underpins Theorem 3.2(i).
  • domain assumption The container results [PV11, Proposition 2.7] and [DV24, Lemma 5.4] on relative amenability and intertwining are used as black boxes throughout Sections 4 and 5.
    These results are cited and not reproved; they are load-bearing for Lemma 5.2 and Theorem 5.1.
  • domain assumption The free group factors L(F_n) satisfy (AO)+ and W*CBAP, hence are stably solid by Proposition 3.7.
    This is a standard fact in the field and is needed for the vertex algebras in Theorem 6.4.

how reviews work

0 comments
Cite this review

Pith. "Pith review of W*-correlations of II$_1$ factors and rigidity of tensor products and graph products." pith.science (2026). https://pith.science/paper/5TBHA3IS

@misc{pith2026250704691,
  author       = {Pith},
  title        = {Pith review of: W*-correlations of II$_1$ factors and rigidity of tensor products and graph products},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5TBHA3IS}},
  note         = {Machine review of arXiv:2507.04691}
}
abstract

A variant of Gromov's notion of measure equivalence for groups has been introduced for II$_1$ factors under different names. We propose the terminology of W*-correlated II$_1$ factors. We prove rigidity results up to W*-correlations for tensor products and graph products of II$_1$ factors. As a consequence, we construct the first uncountable family of discrete groups $\Gamma$ that are not von Neumann equivalent, which means that their group von Neumann algebras $L(\Gamma)$ are not W*-correlated, and which implies that these groups are neither measure equivalent, nor have isomorphic or virtually isomorphic group von Neumann algebras.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

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

  1. A continuum of non-measure equivalent groups

    math.GR 2025-12 conditional novelty 7.0 of 10

    A continuum of pairwise non-measure-equivalent finitely generated property (T) groups with all ℓ²-Betti numbers zero is constructed.

  2. Universal C$^{\ast}$-Algebras from Graph Products: Structure and Applications

    math.OA 2025-07 conditional novelty 7.0 of 10

    A new family of ambient C*-algebras around graph products yields universal properties, nuclearity/exactness characterizations, a maximal ideal, and new simplicity and trace-uniqueness criteria for graph product C*-algebras.

Reference graph

Works this paper leans on

37 extracted references · 34 canonical work pages · cited by 2 Pith papers

  1. [1]

    Avsec, Strong solidity of the q -Gaussian Algebras for all -1 < q < 1

    S. Avsec, Strong solidity of the q -Gaussian Algebras for all -1 < q < 1 . Preprint. arXiv:1110.4918 https://arxiv.org/abs/1110.4918

    arXiv

  2. [2]

    Berbec and S

    M. Berbec and S. Vaes, W ^* -superrigidity for group von Neumann algebras of left-right wreath products. Proc. London Math. Soc. 108 (2014), 1116-1152

    work page 2014

  3. [3]

    Berendschot and S

    T. Berendschot and S. Vaes, Measure equivalence embeddings of free groups and free group factors. Ann. Sci. \' E c. Norm. Sup\' e r. 58 (2025), 389-418

    work page 2025

  4. [4]

    Borst, M

    M. Borst, M. Caspers and E. Chen, Rigid graph products. Preprint. arXiv:2408.06171 https://arxiv.org/abs/2408.06171

    arXiv

  5. [5]

    Boutonnet, On solid ergodicity for Gaussian actions

    R. Boutonnet, On solid ergodicity for Gaussian actions. J. Funct. Anal. 263 (2012), 1040-1063

    work page 2012

  6. [6]

    Bo\.zejko and R

    M. Bo\.zejko and R. Speicher, An example of a generalized Brownian motion. Comm. Math. Phys. 137 (1991), 519-531

    work page 1991

  7. [7]

    Brown and N

    N.P. Brown and N. Ozawa, C ^* -algebras and finite-dimensional approximations. Graduate Studies in Mathematics 88 . American Mathematical Society, Providence, 2008

    work page 2008

  8. [8]

    Caspers and P

    M. Caspers and P. Fima, Graph products of operator algebras. J. Noncommut. Geom. 11 (2017), 367-411

    work page 2017

Show all 37 references
  1. [9]

    Caspers, Y

    M. Caspers, Y. Isono and M. Wasilewski, L ^2 -cohomology, derivations, and quantum Markov semi-groups on q -Gaussian algebras. Int. Math. Res. Not. IMRN 9 (2021), 6405-6441

    2021

  2. [10]

    Davis, The geometry and topology of Coxeter groups

    M.W. Davis, The geometry and topology of Coxeter groups. London Math. Soc. Monogr. Ser. 32 , Princeton University Press, Princeton, 2008

    2008

  3. [11]

    C. Ding, S. Kunnawalkam Elayavalli and J. Peterson, Properly proximal von Neumann algebras. Duke Math. J. 172 (2023), 2821-2894

    2023

  4. [12]

    Ding and J

    C. Ding and J. Peterson, Biexact von Neumann algebras. Preprint. arXiv:2309.10161 https://arxiv.org/abs/2309.10161

    arXiv

  5. [13]

    Donvil and S

    M. Donvil and S. Vaes, W ^* -superrigidity for cocycle twisted group von Neumann algebras. Invent. Math. 240 (2025), 193-260

    2025

  6. [14]

    Drimbe, Measure equivalence rigidity via s-malleable deformations

    D. Drimbe, Measure equivalence rigidity via s-malleable deformations. Compos. Math. 159 (2023), 2023-2050

    2023

  7. [15]

    Drimbe, D

    D. Drimbe, D. Hoff, and A. Ioana, Prime II _1 factors arising from irreducible lattices in products of rank one simple Lie groups. J. Reine. Angew. Math. 757 (2019), 197-246

    2019

  8. [16]

    Effros, Property and inner amenability

    E.G. Effros, Property and inner amenability. Proc. Amer. Math. Soc. 47 (1975), 483-486

    1975

  9. [17]

    Gromov, Asymptotic invariants of infinite groups

    M. Gromov, Asymptotic invariants of infinite groups. In Geometric group theory, Vol.\ 2, Sussex, 1991 , London Math. Soc. Lecture Note Ser. 182 , Cambridge University Press, Cambridge, 1993, pp.\ 1-295

    1991

  10. [18]

    Houdayer and Y

    C. Houdayer and Y. Isono, Unique prime factorization and bicentralizer problem for a class of type III factors. Adv. Math. 305 (2017), 402-455

    2017

  11. [19]

    Ioana, Cartan subalgebras of amalgamated free product II _1 factors

    A. Ioana, Cartan subalgebras of amalgamated free product II _1 factors. Ann. Sci. \' E c. Norm. Sup\' e r. (4) 48 (2015), 71-130

    2015

  12. [20]

    Ishan, Von Neumann equivalence and group approximation properties

    I. Ishan, Von Neumann equivalence and group approximation properties. Groups Geom. Dyn. 18 (2024), 737-747

    2024

  13. [21]

    Ishan, J

    I. Ishan, J. Peterson and L. Ruth, Von Neumann equivalence and properly proximal groups. Adv. Math. 438 (2024), paper no.\ 109481

    2024

  14. [22]

    Isono, Examples of factors which have no Cartan subalgebras

    Y. Isono, Examples of factors which have no Cartan subalgebras. Trans. Amer. Math. Soc. 367 (2015), 7917-7937

    2015

  15. [23]

    Isono, Unique prime factorization for infinite tensor product factors

    Y. Isono, Unique prime factorization for infinite tensor product factors. J. Funct. Anal. 276 (2019), 2245-2278

    2019

  16. [24]

    Isono and A

    Y. Isono and A. Marrakchi, Tensor product decompositions and rigidity of full factors. Ann. Sci. \' E c. Norm. Sup\' e r. 55 (2022), 109-139

    2022

  17. [25]

    Monod and Y

    N. Monod and Y. Shalom, Orbit equivalence rigidity and bounded cohomology. Ann. of Math. 164 (2006), 825-878

    2006

  18. [26]

    Nou, Non injectivity of the q -deformed von Neumann algebra

    A. Nou, Non injectivity of the q -deformed von Neumann algebra. Math. Ann. 330 (2004), 17-38

    2004

  19. [27]

    Ozawa, Solid von Neumann algebras

    N. Ozawa, Solid von Neumann algebras. Acta Math. 192 (2004), 111-117

    2004

  20. [28]

    Ozawa, A Kurosh-type theorem for type II _1 factors

    N. Ozawa, A Kurosh-type theorem for type II _1 factors. Int. Math. Res. Not. Art. ID 97560 (2006), 21 pp

    2006

  21. [29]

    Ozawa and S

    N. Ozawa and S. Popa, Some prime factorization results for type II _1 factors. Invent. Math. 156 (2004), 223-234

    2004

  22. [30]

    Ozawa and S

    N. Ozawa and S. Popa, On a class of II _1 factors with at most one Cartan subalgebra. Ann. Math. 172 (2010), 713-749

    2010

  23. [31]

    Popa, Strong rigidity of II _1 factors arising from malleable actions of w -rigid groups, I

    S. Popa, Strong rigidity of II _1 factors arising from malleable actions of w -rigid groups, I. Invent. Math. 165 (2006), 369-408

    2006

  24. [32]

    Popa, Deformation and rigidity for group actions and von Neumann algebras

    S. Popa, Deformation and rigidity for group actions and von Neumann algebras. In International Congress of Mathematicians, vol.\ I , European Mathematical Society (EMS), Z\" u rich, 2007, pp.\ 445-477

    2007

  25. [33]

    Popa and S

    S. Popa and S. Vaes, Unique Cartan decomposition for II _1 factors arising from arbitrary actions of free groups. Acta Math. 212 (2014), 141-198

    2014

  26. [34]

    Popa and S

    S. Popa and S. Vaes, Unique Cartan decomposition for II _1 factors arising from arbitrary actions of hyperbolic groups. J. Reine Angew. Math. 694 (2014), 215-239

    2014

  27. [35]

    Ricard, Factoriality of q -Gaussian von Neumann algebras

    \' E . Ricard, Factoriality of q -Gaussian von Neumann algebras. Comm. Math. Phys. 257 (2005), 659-665

    2005

  28. [36]

    Vaes, An inner amenable group whose von Neumann algebra does not have property Gamma

    S. Vaes, An inner amenable group whose von Neumann algebra does not have property Gamma. Acta Math. 208 (2012), 389-394

    2012

  29. [37]

    Wildschut, Strong solidity of q -Gaussian algebras

    A.G. Wildschut, Strong solidity of q -Gaussian algebras. Master thesis Delft University of Technology (2020). Available here. https://resolver.tudelft.nl/a0170f6a-cf6d-4349-8530-dbd60471bab4

    2020

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.