$C^*$-extreme contractive completely positive maps

Anand O. R; K. Sumesh

$C^*$-extreme contractive completely positive maps

T0 review · 0 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read This paper proves that for finite-dimensional Hilbert spaces, the C*-extreme points of contractive completely positive maps are exactly the maps $\Phi$ for which $\Phi(1)$ is a projection and the nonzero block is a C*-extreme unital…

desk verdict A careful, correct finite-dimensional completion of the Farenick-Zhou program for contractive CP maps via a genuinely new P-C*-convexity framework. read the letter →

arxiv 2412.05008 v3 pith:GHBZXY2J submitted 2024-12-06 math.OA math.FA

Anand O. R , K. Sumesh This is my paper · ORCID

classification math.OAmath.FA MSC 46L0546L0746L30

keywords C*-algebracompletelypositivemapC*-convexityC*-extremepointP-C*-convexitycontractiveKrein-Milmantheorem

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

This paper gives a complete classification of the C*-extreme points of the set of contractive completely positive maps from a unital C*-algebra into $B(\mathcal{H})$, when $\mathcal{H}$ is finite-dimensional. A C*-extreme point here is an extreme point with respect to quantum convex combinations, where the scalar weights are replaced by operators $T_j$ satisfying $\sum_j T_j^*T_j = I$. The paper proves that a contractive map $\Phi$ is C*-extreme exactly when $\Phi(1)$ is a projection and the restriction to the range of $\Phi(1)$ is a C*-extreme unital completely positive map; equivalently, $\Phi$ is a unitary conjugate of a direct sum of nested pure unital completely positive maps padded with a zero block. To reach this, it introduces $P$-C*-convexity, a generalized quantized convexity indexed by a positive operator $P$, and studies the extreme points of the affine slice $\mathrm{CP}^{(P)}(A,B(\mathcal{H})) = \{\Phi : \Phi(1)=P\}$. This generalizes the earlier C*-convexity theory for unital maps and yields Krein-Milman-type theorems asserting that the whole sets are generated, in the bounded-weak topology, by their respective extreme points.

What carries the argument

The machinery is $P$-C*-convexity: with $P \in B(\mathcal{H})_+$, a combination $\sum_j \mathrm{Ad}_{T_j} \Phi_j$ with $\sum_j T_j^* P T_j = P$ is a $P$-C*-convex combination, and $\mathrm{CP}^{(P)}(A,B(\mathcal{H})) = \{\Phi \in \mathrm{CP} : \Phi(1)=P\}$ carries this structure. The load-bearing facts are the abstract characterization (Theorem 3.12): $\Phi$ is $P$-C*-extreme iff every CP map $\Psi \le_{cp} \Phi$ with $\Psi(1) = B^* P B$ for invertible $B$ satisfies $\Psi = \mathrm{Ad}_Z \Phi$ for some invertible $Z$; the invertible-conjugation theorem (Theorem 3.19) transferring $P$-extremality to UCP-extremality; and the block-triangular reduction (Proposition 3.17 and Corollary 3.16) that in finite dimensions strips away the zero block on $\mathrm{range}(P)^\perp$. Together they convert the classification of contractive C*-extreme maps into the known classification of unital C*-extreme maps.

What would settle it

Search small finite-dimensional examples for a C*-extreme contractive map $\Phi$ with $\Phi(1)$ not a projection; the theorem asserts none exists. For instance, with $\mathcal{A}=M_2$ and $\mathcal{H}=\mathbb{C}^3$, any $\Phi$ with $\Phi(1)=\operatorname{diag}(1,\tfrac12,0)$ must admit a proper C*-convex decomposition into two summands not unitarily equivalent to $\Phi$, and Lemma 4.10 gives the explicit decomposition to check. Finding one map for which every such decomposition still forces unitary equivalence would refute Theorem 4.12.

Watch

Extended reading notes

Core claim

The central discovery is that, inside the finite-dimensional contractive set, C*-extremality is a projection phenomenon. Theorem 4.12 proves that $\Phi \in \mathrm{CCP}(A,B(\mathcal{H}))$ is C*-extreme if and only if $P := \Phi(1)$ is a projection and $\Phi$ is a $P$-C*-extreme point of the slice $\mathrm{CP}^{(P)}(A,B(\mathcal{H}))$. By the structural theorem Theorem 3.20, this means $\Phi$ is unitarily equivalent to a block map $\left(\bigoplus_{i,j} \Phi^{\pi_i}_j\right) \oplus 0$ with respect to $\mathcal{H} = \left(\bigoplus_{i,j} \mathcal{H}^i_j\right) \oplus \mathrm{range}(P)^\perp$, where each $\Phi^{\pi_i}_j$ is a pure unital completely positive map forming a nested sequence of compressions of an irreducible representation $\pi_i$. The argument passes through the general theory of $P$-C*-convexity: for invertible $P$, $\Phi$ is $P$-C*-extreme exactly when its normalized version $\widehat{\Phi} = P^{-1/2}\Phi(\cdot)P^{-1/2}$ is a C*-extreme unital map, and the finite-dimensional reduction strips away the zero block to reduce non-invertible $P$ to this case.

Load-bearing premise

The classification rests on finite-dimensionality of the target Hilbert space $\mathcal{H}$; the infinite-dimensional analogue is explicitly left open, and the proof uses finite-dimensionality to obtain block-triangular forms, closed range of $\Phi(1)$, and reduction to invertible $P$.

Editorial extensions

If this is right

In finite dimensions, C*-extreme contractive completely positive maps are automatically linear extreme points of $\mathrm{CCP}(A,B(\mathcal{H}))$.
A Krein-Milman-type theorem holds: when $\mathcal{H}$ is finite-dimensional, the C*-convex hull of the C*-extreme points of $\mathrm{CCP}(A,B(\mathcal{H}))$ is BW-dense in $\mathrm{CCP}(A,B(\mathcal{H}))$.
For commutative $A$, the C*-extreme contractive maps are precisely the $*$-homomorphisms from $A$ into $B(\mathcal{H})$.
For every positive $P$ with finite-dimensional $\mathcal{H}$, the set $\mathrm{CP}^{(P)}(A,B(\mathcal{H}))$ is the BW-closure of the $P$-C*-convex hull of its $P$-C*-extreme points.
When $P$ is invertible, $P$-C*-extremality is equivalent to unital C*-extremality after conjugation, so the contractive classification inherits the nested-compression structure of unital extreme maps.

Reading between the lines

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

The pivotal open step for infinite dimensions is proving that every C*-extreme contractive map has closed range at 1; with that, Lemma 4.10 would force $\Phi(1)$ to be a projection and the padded-unital structure would follow for arbitrary Hilbert spaces.
The same $P$-C*-convexity machinery could classify extremality in other affine slices, such as maps with $\Phi(1)$ equal to a fixed positive contraction or a fixed positive element of a general C*-algebra.
In quantum information terms, the structure says C*-extreme contractive maps act as a unital channel on one subspace and are exactly zero on the complement, a noiseless-subsystem-plus-dark-subspace pattern that could be tested on small matrix examples.
If the finite-dimensional classification is combined with the known structure of C*-extreme unital maps for infinite-dimensional separable settings, a plausible conjecture is that the same union-over-projections formula holds whenever $\Phi(1)$ has closed range.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

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

Worth reading and worth refereeing. The paper completes the finite-dimensional half of a long-running program: it characterizes the C*-extreme points of contractive completely positive maps into B(H) for finite-dimensional H. The key new device is P-C*-convexity, where the usual C*-convex combination condition is weighted by a positive operator P. The main results are Theorem 3.12 (abstract characterization of P-C*-extreme points), Theorem 3.20 (structure as invertible conjugates of nested direct sums of pure UCP maps plus a zero block), and Theorem 4.12, which says that for finite-dimensional H the C*-extreme points of CCP(A,B(H)) are exactly the union over projections P of the P-C*-extreme points of CP(P). That is a clean completion of the Farenick-Zhou classification from the unital to the contractive setting.

The proof strategy adapts known tools, but the adaptation is real and nontrivial, particularly the block-matrix arguments in Section 3 and Lemma 4.10 showing that a C*-extreme point of CCP must have Φ(1) equal to a projection. I also checked the one truly external step in Theorem 4.12, the use of [LoPa81, Prop 26] to promote Φ_j(1)=U_j^*PU_j to unitaries. The stress-test note verifies it directly; the verification is sound, and the conclusion holds. I found no error affecting the central claim.

Soft spots: the finite-dimensional restriction is essential and is honestly flagged in Note 4.14; the infinite-dimensional analogue remains open. That limits the reach of the main theorem, but it is not a flaw in the proof. The paper is also dense and technical; a reader has to work through several auxiliary results before seeing the main structure. The relation between C*-extreme and linear extreme points is handled, but it is not the main event. Citation practice is fine: the authors use standard sources and do not lean on their own prior work.

This paper is for operator algebraists with a serious interest in C*-convexity and extreme points of CP maps. The classification theorem will be a useful reference, and P-C*-convexity might be reusable in other problems. My recommendation: send it to a competent referee, accept after minor revision. The math is correct, the novelty is solid, and the paper is honest about its scope.

Referee Report

0 major / 6 minor

Summary. The paper develops a generalization of C*-convexity called P-C*-convexity, for a positive operator P on a Hilbert space H, and studies the corresponding extreme points in the sets CP(P)(A,B(H)) of completely positive maps with Φ(1)=P. The main contributions are an abstract characterization of P-C*-extreme points (Theorem 3.12), a structure theorem for finite-dimensional H (Theorem 3.20), a Krein-Milman type theorem for CP(P) (Theorem 3.25), and the central application: for finite-dimensional H, the C*-extreme points of the contractive CP maps CCP(A,B(H)) are exactly the union over projections P of the P-C*-extreme points of CP(P)(A,B(H)) (Theorem 4.12). The paper also proves a Krein-Milman theorem for CCP and shows that, in finite dimensions, C*-extreme points of CCP are linear extreme points.

Significance. If the results are correct, the paper gives a complete structural description of C*-extreme contractive completely positive maps into matrix algebras, extending the Farenick-Zhou theory from unital maps to the contractive setting. The main theorem is concrete and checkable, and the proofs are carried out with detailed block-matrix arguments. The authors are also appropriately cautious: Note 4.14 explicitly flags that the infinite-dimensional analogue of the main characterization remains open, and the finite-dimensionality hypotheses in Corollary 3.16, Proposition 3.17, and Lemma 4.10 are made clear. I verified the external projection fact used in the first half of Theorem 4.12, and it is sound. The paper should be of interest to researchers in quantized convexity, completely positive maps, and operator algebras.

minor comments (6)

[Lemma 3.14] The statement uses the decomposition H = H0 ⊕ H0^⊥ with H0 = range(P). This is not valid unless range(P) is closed. The argument and all later finite-dimensional applications go through if H0 is replaced by the closure of range(P); please revise the statement and proof accordingly.
[Theorem 3.7] The stated equivalence is false in the 'only if' direction: if B is a rank-one projection and C is an invertible rotation, A=BC has range(A)=range(B) but ker(A)≠ker(B). The 'if' direction is the one used in the paper, so the theorem should be restated or restricted to positive operators.
[Theorem 3.20] The proof explicitly treats only the case where P is not invertible; the invertible case is not written out. Please add a sentence covering P invertible, since it is needed for the full statement of the theorem.
[Example 4.7(i)] The appeal to Lemma 4.3 is not immediate because the decomposition in the example is a scalar convex combination rather than a C*-convex one. The conclusion s=t follows by comparing the two scalar multiples and using that both summands lie in CCP×; please clarify the argument.
[Theorem 4.12] In the converse direction, the case P=Φ(1)=0 is not addressed. The zero map is indeed in CP(0)C*-ext, but this case should be stated explicitly.
[General presentation] There are several typographical errors, including 'the the' and 'deﬁned with in' in the abstract; the reference [Zhu98] also lists the author as H. Zhuo, which should be corrected.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the main characterization is derived from stated definitions and external theorems, not from its own conclusion.

full rationale

The paper's derivation is self-contained in the sense required for this pass: it fits no parameters to data, and its main structural conclusions do not reduce to its definitions. Theorem 3.12 extends Zhou's abstract characterization, cited as [Zhu98, Theorem 3.1.5] in the corrected form [BhKu, Corollary 2.5]; Theorem 3.19 conjugates the CP(P) problem to the UCP problem via the explicit invertible rescaling \hat\Phi = \Phi(1)^{-1/2}\Phi(\cdot)\Phi(1)^{-1/2}; Theorem 3.20 then imports the Farenick-Zhou classification ([FaZh98, Theorem 2.1]) as an external, non-overlapping structural input. The main theorem 4.12 has two directions: the inclusion from P-C*-extreme points uses the external Loebl-Paulsen/Wu projection lemma, the definition of P-C*-convex combinations, and Lemma 4.11 to upgrade invertible equivalence to unitary equivalence; the reverse uses Lemma 4.10 to show that \Phi(1) is a projection and a block reduction to UCP C*-extreme points. None of these steps assumes the conclusion. The only self-citation, [BDMS23] (which includes the second author), appears in a list of related C*-convexity references and is not load-bearing. Note 4.14 explicitly flags the infinite-dimensional case as open, which is a stated limitation rather than a hidden assumption. No circular step of any of the enumerated kinds is present.

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

No free parameters are fitted to data; the paper's claims are theorems. The central results rely on several external theorems from the C*-convexity literature, which are listed as axioms. No new physical entities are postulated.

assumptions (5)

standard math Arveson's Radon-Nikodym theorem for CP maps (Theorem 2.1): for Phi with minimal Stinespring dilation (K,pi,V), Psi <=cp Phi iff Psi(.) = V* D pi(.) V for a unique positive contraction D in pi(A)'.
Invoked repeatedly, starting in Theorem 3.12, to convert domination by Phi into an operator D on the dilation space.
domain assumption Farenick-Zhou structure theorem (Theorem 2.6): in finite dimensions, UCP C*-extreme points are direct sums of nested pure UCP compressions of pairwise non-equivalent irreducible representations.
Used as the black-box classification of the unital case in Theorems 3.20 and 4.12.
domain assumption Loebl-Paulsen/Wu result (cited as [LoPa81, Proposition 26] and [Wei02] in Theorem 4.12): if P is a projection and P = sum_j T_j* Q_j T_j with sum_j T_j* T_j = I, T_j invertible and Q_j positive contractions, then each Q_j is unitarily equivalent to P.
The key external lemma that converts a C*-convex decomposition at the level of Phi(1) into unitarily equivalent projections; not proved in the paper.
standard math Douglas range factorization and Fillmore-Williams theorem (Theorem 3.7): A = BC with C invertible iff range(A)=range(B) and ker(A)=ker(B).
Used in Lemma 3.2 and Lemma 3.8 to construct invertible operators from range equalities.
standard math Commutant of the spatial tensor product with an irreducible representation: (I_H tensor pi(A))' = B(H) tensor I_K when pi(A)' = CI_K.
Used in Proposition 3.13(i) to identify the compression of the Stinespring dilation; supported by Takesaki [Tak79].

how reviews work

0 comments

Cite this review

Pith. "Pith review of $C^*$-extreme contractive completely positive maps." pith.science (2026). https://pith.science/paper/GHBZXY2J

@misc{pith2026241205008,
  author       = {Pith},
  title        = {Pith review of: $C^*$-extreme contractive completely positive maps},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GHBZXY2J}},
  note         = {Machine review of arXiv:2412.05008}
}

abstract

In this paper we generalize a specific quantized convexity structure of the generalized state space of a $C^*$-algebra and examine the associated extreme points. We introduce the notion of $P$-$C^*$-convex subsets, where $P$ is any positive operator on a Hilbert space $\mathcal{H}$. These subsets are defined with in the set of all completely positive (CP) maps from a unital $C^*$-algebra $\mathcal{A}$ into the algebra $B(\mathcal{H})$ of bounded linear maps on $\mathcal{H}$. In particular, we focus on certain $P$-$C^*$-convex sets, denoted by $\mathrm{CP}^{(P)}(\mathcal{A},B(\mathcal{H}))$, and analyze their extreme points with respect to this new convexity structure. This generalizes the existing notions of $C^*$-convex subsets and $C^*$-extreme points of unital completely positive maps. We significantly extend many of the known results regarding the $C^*$-extreme points of unital completely positive maps into the context of $P$-$C^*$-convex sets we are considering. This includes abstract characterization and structure of $P$-$C^*$-extreme points. Further, using these studies, we completely characterize the $C^*$-extreme points of the $C^*$-convex set of all contractive completely positive maps from $\mathcal{A}$ into $B(\mathcal{H})$, where $\mathcal{H}$ is finite-dimensional. Additionally, we discuss the connection between $P$-$C^*$-extreme points and linear extreme points of these convex sets, as well as Krein-Milman type theorems.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

26 extracted references · 25 canonical work pages

[1]

W. B. Arveson. Subalgebras of C -algebras. Acta Math. , 123:141--224, 1969

work page 1969
[2]

Banerjee, B

T. Banerjee, B. V. Rajarama Bhat, and M. Kumar. C^* -extreme points of positive operator valued measures and unital completely positive maps. Comm. Math. Phys. , 388(3):1235--1280, 2021

work page 2021
[3]

B. V. Rajarama Bhat, R. Devendra, N. Mallick, and K. Sumesh. C^* -extreme points of entanglement breaking maps. Rev. Math. Phys. , 35(3):Paper No. 2350005, 17, 2023

work page 2023
[4]

Balasubramanian and N

S. Balasubramanian and N. Hotwani. C^ -extreme entanglement breaking maps on operator systems. Linear Algebra Appl. , 685:182--213, 2024

work page 2024
[5]

R. Bhatia. Positive definite matrices . Princeton Series in Applied Mathematics. Princeton University Press, Princeton, NJ, 2007

work page 2007
[6]

B. V. Rajarama Bhat and M. Kumar. C^ -extreme maps and nests. J. Funct. Anal. , 282(8):Paper No. 109397, 40, 2022

work page 2022
[7]

M. D. Choi. Completely positive linear maps on complex matrices. Linear Algebra Appl. , 10:285--290, 1975

work page 1975
[8]

J. B. Conway. A course in operator theory , volume 21 of Graduate Studies in Mathematics . American Mathematical Society, Providence, RI, 2000

work page 2000

Show all 26 references

[9]

J. Dixmier. \'etude sur les vari\'et\'es et les op\'erateurs de J ulia, avec quelques applications. Bull. Soc. Math. France , 77:11--101, 1949

1949
[10]

K. R. Davidson and M. Kennedy. Noncommutative choquet theory. arXiv:1905.08436v3 , 2022

1905 arXiv
[11]

R. G. Douglas. On majorization, factorization, and range inclusion of operators on H ilbert space. Proc. Amer. Math. Soc. , 17:413--415, 1966

1966
[12]

E. G. Effros and S. Winkler. Matrix convexity: operator analogues of the bipolar and H ahn- B anach theorems. J. Funct. Anal. , 144(1):117--152, 1997

1997
[13]

D. R. Farenick and P. B. Morenz. C^* -extreme points in the generalised state spaces of a C^* -algebra. Trans. Amer. Math. Soc. , 349(5):1725--1748, 1997

1997
[14]

Fujimoto

I. Fujimoto. CP -duality for C^ - and W^ -algebras. J. Operator Theory , 30(2):201--215, 1993

1993
[15]

P. A. Fillmore and J. P. Williams. On operator ranges. Advances in Math. , 7:254--281, 1971

1971
[16]

D. R. Farenick and H. Zhou. The structure of C^* -extreme points in spaces of completely positive linear maps on C^* -algebras. Proc. Amer. Math. Soc. , 126(5):1467--1477, 1998

1998
[17]

M. C. Gregg. On C^ -extreme maps and -homomorphisms of a commutative C^ -algebra. Integral Equations Operator Theory , 63(3):337--349, 2009

2009
[18]

R. I. Loebl and V. I. Paulsen. Some remarks on C -convexity. Linear Algebra Appl. , 35:63--78, 1981

1981
[19]

G. J. Murphy. C^* -algebras and operator theory . Academic Press, Inc., Boston, MA, 1990

1990
[20]

V. V. Ostapenko. Matrix convexity. Ukra\"in. Mat. Zh. , 47(1):64--69, 1995

1995
[21]

V. I. Paulsen. Completely bounded maps and operator algebras , volume 78 of Cambridge Studies in Advanced Mathematics . Cambridge University Press, Cambridge, 2002

2002
[22]

E. St rmer. Positive linear maps of operator algebras. Acta Math. , 110:233--278, 1963

1963
[23]

W. F. Stinespring. Positive functions on C^* -algebras. Proc. Amer. Math. Soc. , 6:211--216, 1955

1955
[24]

Takesaki

Ma. Takesaki. Theory of operator algebras. I . Springer-Verlag, New York-Heidelberg, 1979

1979
[25]

W. Wu. C^* -extreme points in W^* -algebras. Acta Math. Sinica (Chinese Ser.) , 45(3):455--460, 2002

2002
[26]

H. Zhuo. C*-extreme points in spaces of completely positive maps . ProQuest LLC, Ann Arbor, MI, 1998. Thesis (Ph.D.)--The University of Regina (Canada)

1998

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.

[1] [1]

W. B. Arveson. Subalgebras of C -algebras. Acta Math. , 123:141--224, 1969

work page 1969

[2] [2]

Banerjee, B

T. Banerjee, B. V. Rajarama Bhat, and M. Kumar. C^* -extreme points of positive operator valued measures and unital completely positive maps. Comm. Math. Phys. , 388(3):1235--1280, 2021

work page 2021

[3] [3]

B. V. Rajarama Bhat, R. Devendra, N. Mallick, and K. Sumesh. C^* -extreme points of entanglement breaking maps. Rev. Math. Phys. , 35(3):Paper No. 2350005, 17, 2023

work page 2023

[4] [4]

Balasubramanian and N

S. Balasubramanian and N. Hotwani. C^ -extreme entanglement breaking maps on operator systems. Linear Algebra Appl. , 685:182--213, 2024

work page 2024

[5] [5]

R. Bhatia. Positive definite matrices . Princeton Series in Applied Mathematics. Princeton University Press, Princeton, NJ, 2007

work page 2007

[6] [6]

B. V. Rajarama Bhat and M. Kumar. C^ -extreme maps and nests. J. Funct. Anal. , 282(8):Paper No. 109397, 40, 2022

work page 2022

[7] [7]

M. D. Choi. Completely positive linear maps on complex matrices. Linear Algebra Appl. , 10:285--290, 1975

work page 1975

[8] [8]

J. B. Conway. A course in operator theory , volume 21 of Graduate Studies in Mathematics . American Mathematical Society, Providence, RI, 2000

work page 2000