Trace-norm rigidity for reduced products of unitary groups and matrix algebras

Andreas Thom; Ben De Bondt

REVIEW 2 major objections 5 minor 24 references

Trace-norm rigidity for reduced products of unitary groups and matrix algebras

T0 review · 2 major / 5 minor · reviewed 2026-08-01 · deepseek-v4-flash

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

Pith's one-line read Under OCA + MA_ℵ1(σ-linked), every isomorphism of tracial unitary reduced products is coordinatewise.

desk verdict Solid, technically rich paper: the unitary-group analogue of [6] is proven with real new inputs, but the forcing-axiom rigidity hinges on a same-author lifting theorem whose application in Theorem 5.9 needs explicit verification. read the letter →

arxiv 2607.19556 v1 pith:MDVSL624 submitted 2026-07-21 math.OA math.GRmath.LO

Ben De Bondt , Andreas Thom This is my paper · ORCID

classification math.OAmath.GRmath.LO MSC 46L0546L1003E5022C05

keywords metricreducedproductsunitarygroupstracenormrigiditycoordinaterecognitionproduct-formisomorphismsOpenColoringAxiommatrixalgebras

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 prove that the abstract group structure of tracial reduced products of finite unitary groups is rigid: any group isomorphism between two such products must come from an almost permutation of coordinates and coordinatewise automorphisms, after identifying asymptotically equivalent dimensions. It establishes the same kind of rigidity for reduced products of matrix algebras, even for center-preserving *-homomorphisms. A sympathetic reader should care because this is a full classification of automorphisms for a natural family of metric ultraproducts, extending earlier results from symmetric groups to unitary groups and matrix algebras. The route is to reduce global rigidity to two local properties: coordinate recognition and product-form rigidity, then prove those properties using new stability and classification results for finite unitary groups in the trace norm.

What carries the argument

The central mechanism is Lemma 1.2, a template that derives full rigidity from two ingredients: coordinate recognition (CR), meaning an isomorphism transports coordinate-restriction relations via an automorphism of the Stone–Čech boundary, and product-form rigidity (PF), meaning every product-form isomorphism is coordinatewise automorphic up to dimension identification. The lemma invokes a metric lifting theorem, quoted from the paper's references, that under OCA + MA_ℵ1(σ-linked) turns a coordinate-fixing isomorphism of reduced products of separable bounded-diameter metric groups into a map of product form. The genuinely new inputs are: a stability theorem (Theorem 2.5) showing that quasi-L

What would settle it

The most direct falsifier would be a counterexample to the metric lifting theorem: construct a coordinate-fixing isomorphism between two reduced products of separable bounded-diameter metric spaces (e.g., finite metric spaces with diameter 1) that is not of product form, even assuming OCA + MA_ℵ1(σ-linked). Alternatively, under ZFC alone one could try to build a non-trivial automorphism of U_HS[(k_n)] that fixes coordinate equivalence relations, which would contradict the conclusion of Theorem 4.5 if it exists.

Watch

Extended reading notes

Core claim

The central claim is Theorem 4.5: assuming OCA + MA_ℵ1(σ-linked), every group isomorphism φ : U_HS[(k_n)] → U_HS[(l_n)] is trivial. Concretely, there is an almost permutation f of the natural numbers with l_n/k_{f(n)} → 1, and coordinatewise automorphisms α_n of U(l_n), such that φ sends [u_n]_n to [α_n(v_n)]_n, where [v_n]_n is the canonical identification of [u_{f(n)}]_n after adjusting dimensions. The parallel Theorem 5.1 for matrix algebras says every unital *-isomorphism of tracial reduced products is inner up to the same coordinate permutation and asymptotically equivalent dimension identification. More generally, Theorem 5.9 classifies center-preserving unital *-homomorphisms between

Load-bearing premise

The load-bearing premise is the metric lifting theorem cited from the literature: under OCA + MA_ℵ1(σ-linked), every coordinate-fixing isomorphism between reduced products of separable bounded-diameter metric spaces is of product form; if this theorem failed, the main rigidity results would not follow.

Editorial extensions

If this is right

If the central claim is correct, the abstract group structure of U_HS[(k_n)] determines the dimension sequence (k_n) up to asymptotic equivalence and the coordinate structure up to an almost permutation, so automorphism groups are exactly the semidirect products of almost permutations with coordinatewise unitary automorphisms.
For matrix algebras, the analogous rigidity means that the tracial reduced product C*-algebra M_HS^∞[(k_n)] has only inner automorphisms up to coordinate permutation and dimension identification, strengthening operator-norm rigidity results to trace norm.
The product-form rigidity results (Proposition 4.3 and Proposition 5.2) hold in ZFC and show that even without forcing axioms, any isomorphism that respects coordinates is already trivial; the forcing axioms are only needed to lift an arbitrary isomorphism to product form.
The classification of almost surjective homomorphisms U(n)→U(m) (Theorem 3.2) is a standalone structural result about finite unitary groups that can be used in other contexts where approximate surjectivity appears.
For center-preserving *-homomorphisms between tracial reduced matrix algebras, the classification (Theorem 5.9) predicts a clean dichotomy: the map is either essentially invisible on a large central set or essentially a direct sum of finite-dimensional amplifications along a finite-to-one coordinate map.

Reading between the lines

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

The paper leaves implicit that the same template may apply to other families of metric groups with suitable stability and coordinate-recognition properties, e.g., finite orthogonal or symplectic groups, as long as the analogues of Theorem 2.5 and Theorem 3.2 hold.
A concrete extension worth testing: whether the coordinate-recognition theorem (Theorem 4.4) can be proved without the uniform bounded normal generation assumption via a different group-theoretic definition of trace, which would remove the need for Lemma 4.12.
If the metric lifting theorem from the references were to fail in some edge case, the main rigidity theorems would collapse; one could test the lifting theorem directly on reduced products of compact Lie groups with nontrivial fundamental groups, where the metric structure is slightly more complicated.
The paper's use of OCA + MA_ℵ1(σ-linked) is probably essential: the introduction notes that under CH, analogous symmetric-group reduced products have wild automorphisms, so one should expect non-trivial isomorphisms of unitary reduced products under CH as well.

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

This is a serious paper, likely correct in its ZFC parts, and the forcing-axiom rigidity is a natural and well-executed continuation of the authors' symmetric-group work. The new inputs are real: Section 2's compact-group stability in normalized Hilbert-Schmidt norm, Theorem 3.2's classification of almost surjective homomorphisms U(n)->U(m), and the coordinate-recognition machinery using bounded normal generation and trace definability. Proposition 4.3 and Theorem 4.4 give clean ZFC results; Theorem 4.5 then follows the template of Lemma 1.2.

Where I would push back: the load-bearing dependency is [7, Theorem 2.3], the metric lifting theorem. Lemma 1.2 uses it for coordinate-fixing isomorphisms. But Theorem 5.9 applies it to a coordinate-fixing ∗-homomorphism that is not shown to be surjective. The stress-test note is right to flag this. Either the lifting theorem is strong enough to cover homomorphisms, in which case the paper should say so explicitly and give the precise statement, or Theorem 5.9 needs a different argument. This is not a fatal flaw by itself -- the ZFC sections are independent of it -- but the full rigidity results are conditional on exactly this point.

The other soft spot is the dependence on [2] for Theorem 5.4, used in Theorems 5.5 and 5.9. That is a same-author preprint; the authors are transparent about it, but an independent check of Theorem 5.4 would be appropriate before relying on it. Minor: in the proof of Proposition 4.14(2), the symbols v', w' are reused for target-side involutions and then for preimages under φ. Confusing but fixable.

Overall, I did not find a fatal error in the ZFC arguments. The structure is coherent, the exposition is careful, and the authors are honest about what is new and what is adaptation. This paper is for people working on metric reduced products, Ulam stability, and the model theory of C*-algebras. It deserves a serious referee. My recommendation: send to peer review, and ask the referee to verify the application of [7] in Theorem 5.9 and the validity of Theorem 5.4 from [2].

Referee Report

2 major / 5 minor

Summary. The paper studies metric reduced products of finite-dimensional unitary groups and matrix algebras with respect to the normalized Hilbert–Schmidt trace norm. The main ZFC results are a compact-group Ulam stability theorem in the trace norm (Thm 2.5), a classification of almost surjective continuous homomorphisms U(n)→U(m) (Thm 3.2), and coordinate-recognition theorems (Thm 4.4 for unitary groups, Thm 5.7 for matrix algebras). Under OCA + MA_ℵ1(σ-linked), the authors prove that every group isomorphism between such reduced unitary products is trivial (Thm 4.5), i.e., induced by an almost permutation and coordinatewise automorphisms. For reduced matrix algebras they obtain rigidity for unital *-isomorphisms (Thm 5.1), a structure theorem for product-form *-homomorphisms (Thm 5.5), and a decomposition theorem for center-preserving *-homomorphisms (Thm 5.9). The framework is an adaptation of the authors' earlier symmetric-group template (Lemma 1.2), with the metric lifting theorem [7] as the forcing-axiom engine.

Significance. If the results are correct, they provide a complete classification of automorphism groups of these 'hyperlinear' reduced products under standard forcing axioms, paralleling the symmetric-group results in [6] and going beyond the operator-norm rigidity of Farah et al. The ZFC sections are technically substantial: the compact-group stability theorem (Section 2), the almost-surjectivity classification (Section 3), and the coordinate recognition via bounded normal generation and trace definability (Section 4) are new and of independent interest. The paper is unusually transparent about its dependence on [7] and the same-author preprint [2]. The main risk is the application of [7] in Theorem 5.9 to a non-surjective homomorphism, which is not covered by the stated version in Lemma 1.2.

major comments (2)

[§5, proof of Theorem 5.9] The proof applies the metric lifting theorem [7, Theorem 2.3] to the coordinate-fixing homomorphism ψ, which is not assumed to be surjective. In this paper the lifting theorem is stated (Lemma 1.2) only for coordinate-fixing isomorphisms. Please state the exact hypotheses of [7, Theorem 2.3] and verify them for ψ; if the theorem requires surjectivity, an additional argument is needed. As written, Theorem 5.9 is not justified. This is load-bearing because Theorem 5.9 is the final structural theorem for center-preserving *-homomorphisms.
[§4.3, proof of Proposition 4.14(2)] The proof uses v′, w′ for the target-side involutions obtained from Lemma 4.12, then writes 'Let v′ = φ(v), w′ = φ(w)' without introducing v,w in the domain. This makes the proof unreadable. Introduce distinct names (e.g., a,b for the target-side involutions; set v=φ^{-1}(a), w=φ^{-1}(b)) and define the symbol c_1 appearing in 'v′ =_{θ(T)} c_1' and 'w′ =_{θ(T)} c_1'. Since Proposition 4.14 is the core of coordinate recognition, this must be corrected.

minor comments (5)

[§3, Theorem 3.2] In the second alternative of the displayed conclusion, the argument v is missing an overline (or should be replaced by the dual / v^{-1}); as printed the two alternatives are identical.
[§2, Theorem 2.5 proof] The proof is a reduction to the arguments in [4] and states that 'the rest of the proof of [4, Theorem 5.2] applies verbatim'. Since this theorem is central to Corollary 2.6 and thus to Proposition 4.3, please spell out the modifications to [4] in more detail.
[§4.2, proof of Lemma 4.7] The estimate that every y∈U(m_n) lies within distance 4√((m_n−l_n)/m_n) of some ṽ is sketched in a parenthetical. A few more details would help.
[Notation 1.1] The notation (NN)∞ is nonstandard and the superscript infinity is not defined; please clarify.
[References] Reference [22] (Takesaki) appears in the bibliography but I did not find it cited in the text.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the main rigidity theorems are genuine derivations from independently stated prior theorems, not restatements of their inputs.

full rationale

I traced the derivation chain. Section 2 proves the compact-group trace-norm stability theorem from measurable approximation, Pettis/Dixmier, and the argument of [4]; Section 3 proves the almost-surjective classification from representation theory and volume estimates; Section 4 proves coordinate recognition (Theorem 4.4) using bounded normal generation [10] and the paper's own definable-trace lemmas, and proves product-form rigidity (Proposition 4.3) from the paper's own Corollary 2.6 and Theorem 3.2. Lemma 1.2 is a general template whose proof reduces an arbitrary isomorphism to product form via the cited metric lifting theorem [7] and then applies product-form rigidity. Theorem 4.5 instantiate this with properties proved in this paper. No premise is defined in terms of the conclusion, no fitted parameter is relabeled as a prediction, and the paper does not merely rename a known result. The self-citations [6], [7], [10], and [2] are external theorems about metric reduced products, finite unitary groups, or Ulam stability; none assumes Theorem 4.5 or Theorem 5.1. The reliance on [7, Theorem 2.3] is a dependence on prior work, not circularity; a hidden failure of that theorem would be a correctness concern, not a circularity concern.

Assumptions & free parameters 0 free parameters · 11 assumptions · 0 invented entities

No free parameters or invented entities appear; all constants in estimates are proven to exist. The central derivations rely on a mix of standard mathematics and several external theorems, several involving the same authors, plus the explicitly stated forcing axioms OCA + MA_ℵ1(σ-linked).

assumptions (11)

standard math ZFC
All unconditional statements are proved in ZFC; the forcing-axiom statements explicitly add OCA + MA_ℵ1(σ-linked).
domain assumption OCA + MA_ℵ1(σ-linked)
Assumed for the main full-rigidity theorems (Lemma 1.2, Theorems 4.5, 5.1, 5.9). This is an external set-theoretic axiom, consistent relative to ZFC, and is explicitly stated.
standard math Classification of low-dimensional irreducible SU(n)-modules
Used in Corollary 3.3 and Theorem 3.2 to identify standard and dual representations below dimension binom(n,2); cited to [14] and [24].
standard math Bishop–Gromov comparison theorem
Used in Lemmas 3.5 and 3.6 to bound covering numbers of PU(n) and its images; cited to [20].
standard math Pettis theorem / Dixmier theorem on continuous representatives of measurable positive definite functions
Used in Proposition 2.4 to replace measurable positive definite kernels by continuous ones; cited to [21] and [8].
standard math Macdonald volume formula for U(n)
Used in Lemma 3.5 to lower-bound the volume and covering number of PU(n); cited to [14].
standard math Bounded normal generation modulo the center for PU(n) ([10, Theorem 1.1 and Lemma 7.8])
Used in Lemma 4.12 to express elements as bounded products of conjugates of a non-central element; this is a strong external theorem from Dowerk–Thom.
standard math Hadwin–Shulman factorization for homomorphisms from finite abelian groups ([16])
Used as Lemma 4.10 to lift homomorphisms from finite abelian groups through the reduced product.
domain assumption Metric lifting theorem ([7, Theorem 2.3])
The key bridge in Lemma 1.2 and Theorem 5.9: under OCA+MA_ℵ1(σ-linked), coordinate-fixing isomorphisms between reduced products of metric spaces are product-form. This is cited, not proved here.
domain assumption Trace-norm Ulam stability for matrix algebras ([2, Theorem 3.5])
Imported as Theorem 5.4 to prove the product-form structure Theorem 5.5 and used in Theorem 5.9. It is a same-author preprint and a load-bearing external input.
domain assumption Structure of Boolean algebra endomorphisms of P(ω)/Fin under OCA+MA ([5, Theorem 4.7])
Used in Theorem 5.9 to decompose center-preserving homomorphisms into a large-kernel part and a finite-to-one coordinate part.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Trace-norm rigidity for reduced products of unitary groups and matrix algebras." pith.science (2026). https://pith.science/paper/MDVSL624

@misc{pith2026260719556,
  author       = {Pith},
  title        = {Pith review of: Trace-norm rigidity for reduced products of unitary groups and matrix algebras},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MDVSL624}},
  note         = {Machine review of arXiv:2607.19556}
}

abstract

We study homomorphisms, with a focus on isomorphisms, between the tracial metric reduced products of finite dimensional unitary groups and of matrix algebras. A variant of Ulam stability for unitary groups and a classification of the almost surjective continuous homomorphisms between finite dimensional unitary groups are proved and then used to show that all isomorphisms of product form of these tracial reduced products are induced by almost permutations of the coordinates and coordinatewise application of automorphisms. We prove coordinate recognition for these reduced products and obtain under set theoretic assumptions rigidity and classification results for their full automorphism groups. For tracial reduced matrix algebras we obtain such rigidity result in the more general context of center-preserving $*$-homomorphisms.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

24 extracted references · 4 linked inside Pith

[6]

arXiv:2412.10802.↑1, 1, 1

Ben De Bondt and Andreas Thom,On automorphism groups of metric reduced products of symmetric groups, 2024. arXiv:2412.10802.↑1, 1, 1

arXiv 2024
[2]

Vadim Alekseev and Andreas Thom,The hyperfiniteII 1-factor is Ulam stable(2026), available at 2606.07369.↑2, 5, 5, 5.4

arXiv 2026
[7]

Mathématique 363(2025), 415–424.↑1, 5

Ben De Bondt and Alessandro Vignati,A metric lifting theorem, Comptes Rendus. Mathématique 363(2025), 415–424.↑1, 5

2025
[1]

Vadim Alekseev, Ilijas Farah, and Andreas Thom,Ulam stability for classes of nuclear C*-algebras (2026), available at2606.03757.↑1

arXiv 2026
[3]

Dubuc and Daniele Mundici,Extending Stone duality to multisets and locally finite MV-algebras, Journal of Pure and Applied Algebra189(2004), no

Roberto Cignoli and Eduardo J. Dubuc and Daniele Mundici,Extending Stone duality to multisets and locally finite MV-algebras, Journal of Pure and Applied Algebra189(2004), no. 1, 37–59.↑4.3

2004
[4]

1, 98–118.↑2, 2

Marcus de Chiffre, Narutaka Ozawa, and Andreas Thom,Operator algebraic approach to inverse and stability theorems for amenable groups, Mathematika65(2019), no. 1, 98–118.↑2, 2

2019
[5]

arXiv:2307.06731.↑5

Ben De Bondt, Ilijas Farah, and Alessandro Vignati,Trivial Isomorphisms between Reduced Products, 2024. arXiv:2307.06731.↑5

arXiv 2024
[8]

15, North-Holland Pub- lishing Co., Amsterdam–New York–Oxford, 1977.↑2 RIGIDITY IN TRACE NORM 27

Jacques Dixmier, C ∗-algebras, North-Holland Mathematical Library, vol. 15, North-Holland Pub- lishing Co., Amsterdam–New York–Oxford, 1977.↑2 RIGIDITY IN TRACE NORM 27

1977

Show all 24 references

[9]

Dowerk,Algebraic and topological properties of unitary groups ofII1 factors, 2015

Philip A. Dowerk,Algebraic and topological properties of unitary groups ofII1 factors, 2015. Dissertation, Universität Leipzig.↑4.9

2015
[10]

Dowerk and Andreas Thom,Bounded normal generation and invariant automatic continuity, Advances in Mathematics346(2019), 124–169.↑1, 4.9, 4.3

Philip A. Dowerk and Andreas Thom,Bounded normal generation and invariant automatic continuity, Advances in Mathematics346(2019), 124–169.↑1, 4.9, 4.3

2019
[11]

3, 489–503.↑1

Ilijas Farah,All automorphisms of all Calkin algebras, Mathematical Research Letters18(2011), no. 3, 489–503.↑1

2011
[12]

2, 619–661.↑1

,All automorphisms of the Calkin algebra are inner, Annals of Mathematics173(2011), no. 2, 619–661.↑1

2011
[13]

,Combinatorial set theory of C*-algebras, Springer Monographs in Mathematics, Springer, Cham, 2019.↑1

2019
[14]

A first course, Graduate Texts in Mathe- matics, vol

William Fulton and Joe Harris,Representation theory. A first course, Graduate Texts in Mathe- matics, vol. 129, Springer, New York, 1991.↑3, 3, 3

1991
[15]

12, 1784–1817.↑2

Timothy Gowers and Omid Hatami,Inverse and stability theorems for approximate representations of finite groups, Sbornik: Mathematics208(2017), no. 12, 1784–1817.↑2

2017
[16]

4, 761–792.↑4.10

Don Hadwin and Tatiana Shulman,Stability of group relations under small Hilbert–Schmidt perturbations, Journal of Functional Analysis275(2018), no. 4, 761–792.↑4.10

2018
[17]

222, Springer, Cham, 2015.↑3

BrianC.Hall,Lie groups, Lie algebras, and representations,Second,GraduateTextsinMathematics, vol. 222, Springer, Cham, 2015.↑3

2015
[18]

Knapp,Lie Groups Beyond an Introduction, Second, Progress in Mathematics, vol

Anthony W. Knapp,Lie Groups Beyond an Introduction, Second, Progress in Mathematics, vol. 140, Birkhäuser, Boston, 2002.↑3

2002
[19]

1, 45–59.↑1

Paul McKenney and Alessandro Vignati,Ulam stability for some classes ofC ∗-algebras, Proceedings of the Royal Society of Edinburgh: Section A Mathematics149(2019), no. 1, 45–59.↑1

2019
[20]

171, Springer, New York, 2016.↑3, 3

Peter Petersen,Riemannian geometry, Third, Graduate Texts in Mathematics, vol. 171, Springer, New York, 2016.↑3, 3

2016
[21]

2, 293–308.↑2, 2

Billy James Pettis,On continuity and openness of homomorphisms in topological groups, Annals of Mathematics52(1950), no. 2, 293–308.↑2, 2

1950
[22]

124, Springer-Verlag, Berlin, 2002.↑

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

2002
[23]

arXiv:2511.03607.↑5

Alessandro Vignati and Deniz Yilmaz,The noncommutative weak extension principle, 2025. arXiv:2511.03607.↑5

2025
[24]

Hermann Weyl,The classical groups, Princeton University Press, Princeton, 1946.↑3, 3 Ben De Bondt, Universität Münster, 48149 Münster, Germany Email address:bdebondt@uni-muenster.de Andreas Thom, TU Dresden, 01062 Dresden, Germany Email address:andreas.thom@tu-dresden.de

1946

Pith tools

Reviewed August 1, 2026 · model on record in the stance chip above.

[1] [6]

arXiv:2412.10802.↑1, 1, 1

Ben De Bondt and Andreas Thom,On automorphism groups of metric reduced products of symmetric groups, 2024. arXiv:2412.10802.↑1, 1, 1

arXiv 2024

[2] [2]

Vadim Alekseev and Andreas Thom,The hyperfiniteII 1-factor is Ulam stable(2026), available at 2606.07369.↑2, 5, 5, 5.4

arXiv 2026

[3] [7]

Mathématique 363(2025), 415–424.↑1, 5

Ben De Bondt and Alessandro Vignati,A metric lifting theorem, Comptes Rendus. Mathématique 363(2025), 415–424.↑1, 5

2025

[4] [1]

Vadim Alekseev, Ilijas Farah, and Andreas Thom,Ulam stability for classes of nuclear C*-algebras (2026), available at2606.03757.↑1

arXiv 2026

[5] [3]

Dubuc and Daniele Mundici,Extending Stone duality to multisets and locally finite MV-algebras, Journal of Pure and Applied Algebra189(2004), no

Roberto Cignoli and Eduardo J. Dubuc and Daniele Mundici,Extending Stone duality to multisets and locally finite MV-algebras, Journal of Pure and Applied Algebra189(2004), no. 1, 37–59.↑4.3

2004

[6] [4]

1, 98–118.↑2, 2

Marcus de Chiffre, Narutaka Ozawa, and Andreas Thom,Operator algebraic approach to inverse and stability theorems for amenable groups, Mathematika65(2019), no. 1, 98–118.↑2, 2

2019

[7] [5]

arXiv:2307.06731.↑5

Ben De Bondt, Ilijas Farah, and Alessandro Vignati,Trivial Isomorphisms between Reduced Products, 2024. arXiv:2307.06731.↑5

arXiv 2024

[8] [8]

15, North-Holland Pub- lishing Co., Amsterdam–New York–Oxford, 1977.↑2 RIGIDITY IN TRACE NORM 27

Jacques Dixmier, C ∗-algebras, North-Holland Mathematical Library, vol. 15, North-Holland Pub- lishing Co., Amsterdam–New York–Oxford, 1977.↑2 RIGIDITY IN TRACE NORM 27

1977