Purifications for Convex Cones

Felix Campidell; Tim Netzer

Purifications for Convex Cones

T0 review · 0 major / 5 minor · reviewed 2026-07-31 · grok-4.5

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

Pith's one-line read Every interior point of an indecomposable homogeneous cone admits a purification built only from cone geometry.

desk verdict Clean geometric take on purification: existence for indecomposable homogeneous cones and a usable uniqueness criterion, with standard examples that land. read the letter →

arxiv 2607.28202 v1 pith:G6T43CMU submitted 2026-07-30 quant-ph math.FA

Felix Campidell , Tim Netzer This is my paper

classification quant-phmath.FA MSC 81P1652A2015A4846L07

keywords purificationconvexconesgeneralizedprobabilistictheoriespositivemapstensorproductsofhomogeneousLorentzmaximallyentangledstate

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 purification principle is usually stated with quantum Hilbert-space structure. This paper strips that away and asks what can be recovered from the geometry of a finite-dimensional proper convex cone alone. It proves that if the cone is indecomposable and the composite cone contains the maximally entangled state, every point in a natural automorphism orbit of a fixed interior unit has a purification; for homogeneous cones this covers every interior point. A matching uniqueness criterion is given in terms of extreme positive maps that fix the unit and local automorphisms of the composite. On the boundary the picture changes: when every proper face is simplicial only pure points can be purified, but non-simplicial faces allow mixed boundary purifications. The results recover ordinary quantum purification for positive-semidefinite cones, yield unique purifications for Lorentz cones, and exhibit both uniqueness and non-uniqueness for k-positive maps, PPT tensors, and polyhedral cones.

What carries the argument

Purifications are extreme rays of Pos(C) ≅ C∨ ¯⊗ C whose evaluation at a fixed interior unit u recovers the target point; the maximally entangled state m (the identity map) seeds the existence argument by transporting under local automorphisms id ⊗ g.

What would settle it

Exhibit an indecomposable homogeneous cone and an interior point with no extreme positive endomorphism sending the chosen unit to that point, or find a cone with all proper faces simplicial that nevertheless purifies a mixed boundary point.

Watch

Extended reading notes

Core claim

If C is an indecomposable proper cone and P is any intermediate tensor cone containing the maximally entangled state m, then every point of the orbit G2(P)·u admits a purification in P. In particular every interior point of an indecomposable homogeneous cone has a purification in the maximal tensor product. Uniqueness up to first-factor local automorphisms holds precisely when the extreme positive maps that fix u and lie in P are themselves those local automorphisms. On the boundary, simplicial faces force any purified nonzero point to be pure.

Load-bearing premise

The whole existence proof needs that automorphisms of an indecomposable cone are extreme among positive maps; if that classical equivalence fails, the construction collapses.

Editorial extensions

If this is right

Every interior point of a Lorentz cone has a unique purification in the maximal tensor product up to first-factor local automorphisms.
Ordinary quantum purifications of positive-definite matrices are recovered as the special case of the PSD cone with the completely-positive composite.
For k-positive maps with k ≥ 2, purifications of the identity remain unique, while the full positive-map cone admits non-unique ones (e.g. the Choi map).
The PPT cone contains no purification of any positive-definite matrix, showing that membership of the maximally entangled state in P is essential.
If every proper face is simplicial, only pure boundary points can be purified; non-simplicial faces allow mixed boundary purifications.

Reading between the lines

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

The geometric criterion suggests a classification program: which GPT state cones are exactly the indecomposable homogeneous ones, and which intermediate tensors keep uniqueness.
Non-uniqueness for the full positive cone on PSD_3 already separates ‘positive’ composites from ‘completely positive’ ones as operationally distinct purification theories.
The boundary dichotomy (simplicial vs non-simplicial faces) gives a concrete test for whether a candidate GPT can purify mixed boundary states without leaving the cone geometry.

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 short, honest convex-geometry paper that isolates how much of the purification principle you can get from the cone alone. The new pieces are Theorem 3.5 / Corollary 3.6 (every point in the G2(P)·u orbit purifies when C is indecomposable and m sits in P; hence every interior point of an indecomposable homogeneous cone purifies in the maximal tensor product) and the uniqueness criterion in Theorem 3.8 (ex_u(Pos(C)) ∩ P ⊆ G1(P)). Proposition 3.4 on simplicial faces and the counterexample 4.8 with a non-pure boundary purification are also cleanly done.

What it does well: the proofs are short and check out from the definitions plus classical extremality (Barker for id extreme iff indecomposable; Loewy–Schneider for Lorentz; Choi for the map example). Lemma 3.3 identifying the marginal with evaluation at u is the right bridge. The examples are useful rather than decorative—PSD recovers the usual story, Lorentz gets uniqueness in the full maximal tensor product (stronger than CP), k-positive collapses to CP uniqueness for k≥2, PPT shows why m ∈ P is necessary, and the square / square⊕R+ pair shows both uniqueness on a polyhedral cone and that non-simplicial faces really do allow non-pure boundary purifications. Citation pattern is appropriate; no circular construction.

Soft spots are minor and local. The square-cone enumeration (“8 automorphisms, 24 extreme rays”) is asserted without the computation; fine for an example but a referee will want a sentence or a reference. Everything routes through Barker’s equivalence, which is classical and correctly scoped to indecomposable cones, so that is not a hidden flaw. The paper does not claim to solve a major open problem outside GPT/cone foundations; it stays inside its lane.

This is for people who work on GPTs, operator systems, or finite-dimensional cone theory and want the geometric content of purification stated cleanly. It deserves a serious referee. I would engage with it and expect it to be citable in that literature.

Referee Report

0 major / 5 minor

Summary. The paper studies purifications of points in a finite-dimensional proper convex cone C using only convex geometry. A purification of c is defined as an extreme ray of the maximal tensor product C∨ ¯⊗ C (equivalently, an extreme positive endomorphism) whose marginal with respect to a fixed interior unit u is c; intermediate cones P containing the maximally entangled state m are also considered. The main existence theorem (Thm. 3.5) shows that if C is indecomposable and P contains m, every point of the local-automorphism orbit G2(P)·u admits a purification in P; as a corollary, every interior point of an indecomposable homogeneous cone purifies in the maximal tensor product (Cor. 3.6). Uniqueness up to first-factor local automorphisms is characterized by the condition ex_u(Pos(C)) ∩ P ⊆ G1(P) (Thm. 3.8). On the boundary, if every proper face is simplicial then only pure points can purify (Prop. 3.4), and a direct-sum counterexample shows the simplicial hypothesis is essential (Ex. 4.8). Applications recover quantum purification, treat Lorentz cones, k-positive maps, the PPT cone, and polyhedral examples illustrating non-uniqueness.

Significance. The work cleanly isolates what the purification principle implies from cone geometry alone, separating it from Hilbert-space structure. The existence and uniqueness criteria are short, self-contained once standard extremality facts (Barker; Loewy–Schneider) are granted, and they apply uniformly to Lorentz cones, PSD cones, and intermediate positivity cones. The examples usefully exhibit both uniqueness (Lorentz, k-positive for k≥2) and non-uniqueness (Choi map on Psd_3; maximal vs. CP tensor product), and the PPT example shows that m ∈ P is essential. This is a solid, well-scoped contribution to the convex-geometric foundations of generalized probabilistic theories; the proofs are elementary and checkable, which is a genuine strength.

minor comments (5)

[Example 4.7] Example 4.7 asserts without proof or external reference that the square cone has “8 automorphisms up to scaling” and that S∨ ¯⊗ S has “24 extreme rays, of which 16 lie in S∨ ⊗ S.” A brief sketch, a citation, or a short computational appendix would make the example independently checkable; as written it is the only non-transparent claim in the applications section.
[Definition 3.1] Definition 3.1 insists that purity is always relative to the maximal tensor product even when the purification is required to lie in a smaller cone P. This is deliberate and used later, but a one-sentence motivation (why membership in ex(P) alone is insufficient) would help readers coming from the GPT literature.
[Examples 4.4–4.5; Preliminaries] Typographical slips: “autormorphisms” (Ex. 4.5), “thathasranktwo” (Ex. 4.4), missing spaces after periods in several places, and the citation style “(3, Lemma 3)” / “(12, Theorem 3.22)” is nonstandard; prefer author–year or numbered references consistently.
[Theorem 3.8] In the uniqueness theorem, the groups G1(P) and G2(P) are defined by preservation of P under one-sided local automorphisms. It would help to note explicitly that for P = C∨ ¯⊗ C one recovers G1 = G2 = Aut(C), so the criterion specializes cleanly to the homogeneous case already treated in Cor. 3.6.
[Section 2] Proposition 2.1 is cited as “(3, Lemma 3)” and Proposition 2.2 as “(12, …)”; adding the author names (Barnum et al.; van Dobben de Bruyn) in the text would improve readability without changing the reference list.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: existence/uniqueness follow from cone definitions plus classical external extremality facts.

full rationale

The central claims (Thm 3.5, Cor 3.6, Thm 3.8, Prop 3.4) are short, self-contained arguments from the paper’s own definitions of purification, marginals, maximally entangled state m, and the groups G1(P)/G2(P). The only load-bearing external input is Barker’s classical equivalence (Prop 2.4: id extreme in Pos(C) iff C indecomposable), which is cited from 1981 literature with no author overlap and is used exactly where the paper assumes indecomposability; it does not encode the purification conclusion. Homogeneity then supplies transitivity on int(C) for Cor 3.6. Uniqueness is an if-and-only-if criterion stated in terms of ex_u(Pos(C)) ∩ P, not an imported uniqueness theorem. Examples (PSD, Lorentz, k-positive, PPT, square cone) illustrate rather than force the general theorems. There are no fitted parameters, no self-citation loops, and no renaming of a known empirical pattern. Derivation chain is independent and non-circular.

Assumptions & free parameters 0 free parameters · 6 assumptions · 2 invented entities

The paper is pure finite-dimensional convex geometry. It inherits standard cone, dual, and tensor-product machinery and one classical extremality equivalence (Barker). No numerical free parameters. The only paper-specific invention is the operational definition of purification as an extreme element of the maximal tensor product with prescribed marginal, plus the auxiliary groups G1(P), G2(P).

assumptions (6)

standard math Finite-dimensional proper convex cones are closed, full-dimensional, and salient; duals and Aut(C) behave as usual.
Section 2 setup; background for all statements.
standard math C∨ ¯⊗ D = Pos(C,D) under the standard identification V' ⊗ W ≅ Lin(V,W).
Used throughout to equate purifications with extreme positive maps (Lemma 3.3).
standard math id ∈ ex(Pos(C)) ⇔ Aut(C) ⊆ ex(Pos(C)) ⇔ C indecomposable (Barker, Thm 2.B.2).
Proposition 2.4; load-bearing for existence of purifications via automorphisms in Thm 3.5.
standard math Extreme rays of the minimal tensor product are pure tensors of extremes; product criterion for pure marginals (Props 2.1–2.2, Cor 2.3).
Cited from Barnum et al. and van Dobben de Bruyn; used for boundary purity arguments.
ad hoc to paper A purification of c is an element of ex(C∨ ¯⊗ C) with marginal c (extremality always relative to the maximal tensor product).
Definition 3.1; the paper’s chosen geometric stand-in for the operational purification principle.
domain assumption Intermediate composite cones P may be any convex cone between minimal and maximal tensor products that contains m when existence is claimed.
GPT modeling choice (Section 1 and Thm 3.5); required so that (id ⊗ g)(m) stays in P.

invented entities (2)

Purification of a cone point (extreme positive map h with h(u)=c) independent evidence
purpose: Translate the operational purification principle into pure cone geometry.
Definition 3.1; not a physical object but a mathematical stand-in. Independent checks are the recovery of quantum purification (Ex 4.1) and Lorentz uniqueness (Thm 4.2).
Groups G1(P) and G2(P) of local automorphisms preserving P
purpose: State existence orbits and uniqueness up to local automorphisms cleanly.
Introduced in Thm 3.5 and Thm 3.8; bookkeeping devices, not new physical entities.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Purifications for Convex Cones." pith.science (2026). https://pith.science/paper/G6T43CMU

@misc{pith2026260728202,
  author       = {Pith},
  title        = {Pith review of: Purifications for Convex Cones},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/G6T43CMU}},
  note         = {Machine review of arXiv:2607.28202}
}

abstract

Motivated by the importance of the purification principle in quantum theory and generalized probabilistic theories, we study purifications using only the geometry of a finite-dimensional proper convex cone. We prove an existence theorem for indecomposable cones and intermediate tensor cones containing the maximally entangled state; in particular, every interior point of an indecomposable homogeneous cone admits a purification. This applies to Lorentz cones, for example. We also give a criterion for uniqueness up to local automorphisms. On the boundary, we show that if every proper face of $C$ is simplicial, then only pure points can admit purifications, and we demonstrate that this conclusion fails in the presence of non-simplicial faces. Examples involving positive semidefinite cones, Lorentz cones, $k$-positive maps, PPT tensors, and polyhedral cones illustrate both existence and non-uniqueness phenomena.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

13 extracted references · 4 canonical work pages

[1]

Szarek.Alice and Bob Meet Banach: The Interface of Asymptotic Geometric Analysis and Quantum Information Theory, volume 223 ofMathe- matical Surveys and Monographs

Guillaume Aubrun and Stanisław J. Szarek.Alice and Bob Meet Banach: The Interface of Asymptotic Geometric Analysis and Quantum Information Theory, volume 223 ofMathe- matical Surveys and Monographs. American Mathematical Society, Providence, RI, 2017. doi:10.1090/surv/223. 7

doi:10.1090/surv/223 2017
[2]

Theory of cones.Linear Algebra and its Applications, 39:263–291, 1981

George Phillip Barker. Theory of cones.Linear Algebra and its Applications, 39:263–291, 1981. doi:10.1016/0024-3795(81)90310-4. 3 Purifications for Convex Cones9

work page doi:10.1016/0024-3795(81)90310-4 1981
[3]

Cloning and broadcasting in generic probabilistic models, 2006

Howard Barnum, Jonathan Barrett, Matthew Leifer, and Alexander Wilce. Cloning and broadcasting in generic probabilistic models, 2006. arXiv:quant-ph/0611295. 3

arXiv 2006
[4]

Information processing in generalized probabilistic theories.Physical Review A, 75(3):032304, 2007

Jonathan Barrett. Information processing in generalized probabilistic theories.Physical Review A, 75(3):032304, 2007. 10.1103/PhysRevA.75.032304. 1

doi:10.1103/physreva.75.032304 2007
[5]

Probabilistic theories with purification.Physical Review A, 81(6):062348, 2010

Giulio Chiribella, Giacomo Mauro D’Ariano, and Paolo Perinotti. Probabilistic theories with purification.Physical Review A, 81(6):062348, 2010. 10.1103/PhysRevA.81.062348. 2

work page doi:10.1103/physreva.81.062348 2010
[6]

Informational derivation of quantum theory.Physical Review A, 84(1):012311, 2011

Giulio Chiribella, Giacomo Mauro D’Ariano, and Paolo Perinotti. Informational derivation of quantum theory.Physical Review A, 84(1):012311, 2011. 10.1103/PhysRevA.84.012311. 2

doi:10.1103/physreva.84.012311 2011
[7]

Positivesemidefinitebiquadraticforms.LinearAlgebraanditsApplications, 12(2):95–100, 1975

Man-DuenChoi. Positivesemidefinitebiquadraticforms.LinearAlgebraanditsApplications, 12(2):95–100, 1975. doi:10.1016/0024-3795(75)90058-0. 7

work page doi:10.1016/0024-3795(75)90058-0 1975
[8]

Notes on extremality of the Choi map.Linear Algebra and its Applications, 439 (10):3156–3165, 2013

Kil-Chan Ha. Notes on extremality of the Choi map.Linear Algebra and its Applications, 439 (10):3156–3165, 2013. doi:10.1016/j.laa.2013.09.011. 7

work page doi:10.1016/j.laa.2013.09.011 2013

Show all 13 references

[9]

Positive operators on then-dimensional ice cream cone.JournalofMathematicalAnalysisandApplications,49(2):375–392,1975

Raphael Loewy and Hans Schneider. Positive operators on then-dimensional ice cream cone.JournalofMathematicalAnalysisandApplications,49(2):375–392,1975. doi:10.1016/0022- 247X(75)90186-9. 6

1975 doi
[10]

On extremal positive maps acting between type I factors.Banach Center Publications, 89(1):201–221, 2010

Marcin Marciniak. On extremal positive maps acting between type I factors.Banach Center Publications, 89(1):201–221, 2010. publisher page. 7

2010
[11]

General probabilistic theories: An introduction.Physics Reports, 1033:1–64,

Martin Plávala. General probabilistic theories: An introduction.Physics Reports, 1033:1–64,
[12]

Tensor products of convex cones, 2020

Josse van Dobben de Bruyn. Tensor products of convex cones, 2020. arXiv:2009.11843. 3

2020 arXiv
[2023]

10.1016/j.physrep.2023.09.001. 1

2023 doi

Pith tools

Reviewed July 31, 2026 · model on record in the stance chip above.

[1] [1]

Szarek.Alice and Bob Meet Banach: The Interface of Asymptotic Geometric Analysis and Quantum Information Theory, volume 223 ofMathe- matical Surveys and Monographs

Guillaume Aubrun and Stanisław J. Szarek.Alice and Bob Meet Banach: The Interface of Asymptotic Geometric Analysis and Quantum Information Theory, volume 223 ofMathe- matical Surveys and Monographs. American Mathematical Society, Providence, RI, 2017. doi:10.1090/surv/223. 7

doi:10.1090/surv/223 2017

[2] [2]

Theory of cones.Linear Algebra and its Applications, 39:263–291, 1981

George Phillip Barker. Theory of cones.Linear Algebra and its Applications, 39:263–291, 1981. doi:10.1016/0024-3795(81)90310-4. 3 Purifications for Convex Cones9

work page doi:10.1016/0024-3795(81)90310-4 1981

[3] [3]

Cloning and broadcasting in generic probabilistic models, 2006

Howard Barnum, Jonathan Barrett, Matthew Leifer, and Alexander Wilce. Cloning and broadcasting in generic probabilistic models, 2006. arXiv:quant-ph/0611295. 3

arXiv 2006

[4] [4]

Information processing in generalized probabilistic theories.Physical Review A, 75(3):032304, 2007

Jonathan Barrett. Information processing in generalized probabilistic theories.Physical Review A, 75(3):032304, 2007. 10.1103/PhysRevA.75.032304. 1

doi:10.1103/physreva.75.032304 2007

[5] [5]

Probabilistic theories with purification.Physical Review A, 81(6):062348, 2010

Giulio Chiribella, Giacomo Mauro D’Ariano, and Paolo Perinotti. Probabilistic theories with purification.Physical Review A, 81(6):062348, 2010. 10.1103/PhysRevA.81.062348. 2

work page doi:10.1103/physreva.81.062348 2010

[6] [6]

Informational derivation of quantum theory.Physical Review A, 84(1):012311, 2011

Giulio Chiribella, Giacomo Mauro D’Ariano, and Paolo Perinotti. Informational derivation of quantum theory.Physical Review A, 84(1):012311, 2011. 10.1103/PhysRevA.84.012311. 2

doi:10.1103/physreva.84.012311 2011

[7] [7]

Positivesemidefinitebiquadraticforms.LinearAlgebraanditsApplications, 12(2):95–100, 1975

Man-DuenChoi. Positivesemidefinitebiquadraticforms.LinearAlgebraanditsApplications, 12(2):95–100, 1975. doi:10.1016/0024-3795(75)90058-0. 7

work page doi:10.1016/0024-3795(75)90058-0 1975

[8] [8]

Notes on extremality of the Choi map.Linear Algebra and its Applications, 439 (10):3156–3165, 2013

Kil-Chan Ha. Notes on extremality of the Choi map.Linear Algebra and its Applications, 439 (10):3156–3165, 2013. doi:10.1016/j.laa.2013.09.011. 7

work page doi:10.1016/j.laa.2013.09.011 2013