Pith. sign in

REVIEW 2 major objections 4 minor 1 cited by

Thermal operations from informational equilibrium

T0 review · 2 major / 4 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 Thermal operations are exactly the quantum channels that admit an equilibrating dilation: a unitary interaction with an environment left invariant whenever the system starts in equilibrium, provided the fixed point has full rank.

desk verdict Clean new characterization of thermal operations via equilibrating dilations; the advertised strict hierarchy is partly conditional on MIP*=RE, but the paper is honest about that and the core result is solid. read the letter →

arxiv 2507.16637 v1 pith:V5QP5QID submitted 2025-07-22 quant-ph

Seok Hyung Lie , Jeongrak Son , Paul Boes , Nelly H.Y. Ng , Henrik Wilming This is my paper · ORCID
classification quant-ph PACS 03.65.Ta
keywords thermaloperationsequilibratingdilationcatalyticchannelsGibbs-preservingmapsdoubly-stochasticfactorizableConnesembeddingproblemdual-unitarycircuits
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 argues that the standard allowed operations of quantum thermodynamics—thermal operations—are nothing more than the channels that admit an 'equilibrating dilation': a unitary coupling to an environment whose reduced state is left unchanged whenever the system is prepared in its equilibrium state. The authors prove that if a channel has such a dilation with respect to a full-rank fixed point, then the unitary must commute with the joint state and the channel must be time-translation covariant; interpreting full-rank states as Gibbs states, this is exactly the defining property of thermal operations. Consequently, any candidate thermalization channel that is Gibbs-preserving but not a thermal operation inevitably disturbs the environment in every dilation, which means the environment must contain non-equilibrium resources. The paper extends the same idea to 'catalytic channels', where the environment is invariant for every input state, and proves a strict hierarchy of doubly-stochastic channels in the fully degenerate case. It also shows that thermal operations already capture all robust catalytic advantages, so no catalyst is needed.

What carries the argument

The load-bearing object is the 'equilibrating dilation' (Definition 1): a dilation $(U,\omega_B)$ of a channel $T$ on system $A$ such that $T(\rho)=\operatorname{Tr}_B(U\rho\otimes\omega_B U^\dagger)$ and the environment is invariant when the system sits in the fixed point $\omega_A$, i.e. $\operatorname{Tr}_A(U\omega_A\otimes\omega_B U^\dagger)=\omega_B$. Lemma 3—the 'informational zeroth law'—shows that equilibrium already forces global stationarity, $U\omega_A\otimes\omega_B U^\dagger = \omega_A\otimes\omega_B$, from which Proposition 4 derives the commutation and covariance properties that make the channel a thermal operation. The second main object is the 'catalytic dilation' (Definition 2), where the environment is invariant for every input state; Proposition 10 characterizes these dilations by the partial transpose $U^{\top_A}$ being unitary, which connects catalytic channels to dual-unitary tensors. The strict inclusions in the doubly-stochastic hierarchy are then proven by extremality arguments (Lemma 13) and Schur multipliers (Proposition 15).

What would settle it

Take a known enhanced thermal operation that is not a thermal operation and numerically search over all dilations $(U,\omega_B)$ of the channel for one where $\operatorname{Tr}_A(U\omega_A\otimes\omega_B U^\dagger)=\omega_B$ with a full-rank Gibbs fixed point; the paper's Eq. (9) claims no such dilation exists, so finding one would refute the central characterization.

Watch

Extended reading notes

Core claim

The central claim is an exact characterization: a quantum channel is a thermal operation if and only if it admits an equilibrating dilation with respect to a full-rank fixed point $\omega_A = e^{-\beta H_A}/Z_A$, meaning there exist a unitary $U$ and environment state $\omega_B$ such that $T(\rho)=\operatorname{Tr}_B(U\rho\otimes\omega_B U^\dagger)$ for all $\rho$ and $operatorname{Tr}_A(U\omega_A\otimes\omega_B U^\dagger)=\omega_B$. Proposition 4 derives from this equilibrium condition that $[U,\omega_A\otimes\omega_B]=0$ and that $T$ is covariant under the modular flow $\rho\mapsto \omega_A^{it}\rho\omega_A^{-it}$; when $\omega_A$ is full rank it can be read as a Gibbs state, so these are precisely time-translation covariance and total energy conservation. Conversely, every thermal operation obviously satisfies the equilibrium condition. A corollary is that the gap between thermal operations and looser classes such as enhanced thermal operations or Gibbs-preserving maps is exactly the failure of equilibrium preservation: for any dilation of such a channel, the environment state must change, even when the system is at equilibrium. For the fully degenerate case the paper determines the structure of doubly-stochastic channels, proving the strict chain $\mathrm{MU}\subsetneq\mathrm{CAT}\subsetneq\mathrm{EQ}\cap\mathrm{DS}=\text{strongly factorizable}\subsetneq F\subsetneq \mathrm{DS}$, and links catalytic channels to dual-unitary circuits.

Load-bearing premise

The proof that thermal operations equal equilibrating channels assumes the fixed point is a full-rank state so that it can be written as a Gibbs state, and the strict inclusion of factorizable channels in the hierarchy rests on the Connes embedding problem having a negative resolution, which the paper notes is not yet fully peer-reviewed.

Editorial extensions

If this is right

  • Thermal operations are the only channels whose environment can stay in equilibrium with a full-rank fixed point; every Gibbs-preserving non-thermal channel has dilations that must disturb the environment, so non-equilibrium resources are unavoidable in any implementation.
  • Enhanced thermal operations and general Gibbs-preserving maps are strictly larger than thermal operations, and the difference is exactly this environmental disturbance.
  • Among doubly-stochastic channels, mixed-unitary, catalytic, equilibrating, and factorizable channels form a strict hierarchy; in particular catalytic channels are not all mixed unitary, and not every doubly-stochastic channel is catalytic.
  • Thermal operations gain nothing from robust catalysis: every robust catalytic thermal operation can be implemented without a catalyst.
  • Catalytic channels correspond to dual-unitary circuits and catalytic unitaries, linking thermodynamic channel structure to exactly solvable circuit models.

Reading between the lines

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

  • If the characterization holds, 'equilibrium' is a more primitive notion than temperature: any full-rank state can serve as the equilibrium state, and the Hamiltonian and inverse temperature emerge from $-\log\omega_A$ rather than being put in by hand; this could extend thermal operations to systems without a pre-specified Hamiltonian.
  • The F-layer of the hierarchy depends on the negative resolution of the Connes embedding problem via MIP*=RE, which the paper itself flags as not yet fully peer-reviewed; if that resolution ever fails, the hierarchy could collapse to MU $\subsetneq$ CAT $\subsetneq$ EQ$\cap$DS = F $\subsetneq$ DS.
  • The duality between catalytic and dual-unitary tensors suggests a two-way transfer: known results about dual-unitary quantum circuits (e.g., integrability and spectral properties) might yield new constraints on catalytic channels, and vice versa.
  • A natural next test is whether the hierarchy survives for non-degenerate Hamiltonians through Gibbs-embedding, a route the paper mentions but leaves open; if it does, the 'environment must change' signature could be used experimentally to certify whether a purported heat bath is truly thermal.
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 / 4 minor

Summary. The paper introduces a purely information-theoretic notion of equilibrium: two local states are in informational equilibrium relative to a unitary if each is preserved as the appropriate marginal of the joint time-evolved state (Eq. (1)). A channel admitting an equilibrating dilation (Definition 1) is shown, via Lemma 3 and Proposition 4, to commute with the fixed-point state in the dilation and to be covariant under the modular flow of that fixed point. Interpreting full-rank fixed points as Gibbs states, the authors conclude that thermal operations are precisely the channels admitting equilibrating dilations with full-rank fixed points. The paper then defines catalytic dilations (Definition 2), proves several structural results about catalytic and factorizable channels, and advertises a strict hierarchy MU⊊CAT⊊EQ∩DS⊊F⊊DS for doubly-stochastic channels (Eq. (12), Fig. 1). The strict inclusion EQ∩DS⊊F relies on the MIP*=RE resolution of the Connes embedding problem, as the paper itself notes in footnote [40] and in the sentence immediately before Eq. (15). The paper also contains a multipartite equilibrium lemma (Lemma 7), a robust-catalysis result for thermal operations (Lemma 8), and a connection between catalytic unitaries and dual-unitary tensors (Appendix B).

Significance. The central characterization of thermal operations is conceptually attractive and, as far as the proofs in the End Matter show, correct: the argument from unitary invariance of entropy and mutual information to [U, ωA⊗ωB]=0 is elementary and does not presuppose thermodynamics. If the advertised hierarchy is understood with the appropriate caveats, the paper resolves previously open questions by separating mixed-unitary, catalytic, equilibrating, and factorizable channels, and it gives a useful new perspective on the gap between thermal operations and Gibbs-preserving maps. The proofs are largely self-contained, use no fitted parameters, and are honest about the external MIP*=RE dependency. The multipartite lemma and the robust-catalysis result are additional contributions that should be of interest to the quantum-thermodynamics community.

major comments (2)
  1. [Hierarchy of doubly-stochastic channels, Eq. (12), Fig. 1, Eq. (15), footnote [40]] The strict inclusion EQ∩DS⊊F is derived from MIP*=RE, which the paper's footnote [40] explicitly says has no completely peer-reviewed proof. The sentence just before Eq. (15) says 'If true', but Eq. (12) and Fig. 1 state 'all inclusions are strict' as an unconditional result, and the Conclusions repeat that the classes are 'all distinct'. Because [13,14] imply that a positive resolution of the Connes embedding problem would make strongly factorizable and factorizable maps coincide in finite dimensions, this particular strict inclusion would collapse under the opposite external verdict. The paper should mark the inclusion as conditional in Eq. (12), in Fig. 1, and in the Conclusions, or supply a proof independent of [15].
  2. [Appendix C, Lemma 13] In the final branch of the proof, where T_{|ψ⟩}=T for all |ψ⟩, the derivation reaches U^{⊤A}(ρ⊗1)(U^{⊤A})† = T(ρ)⊗1. The text then says this 'is possible only when ρ and T(ρ) have the same spectrum ... i.e. T(ρ)=WρW†.' As written, the unitary W is only shown to exist for each fixed ρ, and the proof does not establish that a unital completely positive channel preserving the spectrum of every density matrix must be a single unitary conjugation. Since Lemma 13 is load-bearing for the strict inclusion CAT⊊EQ∩DS in Eq. (12), this step needs either a complete argument or a precise citation to a theorem on spectrum-preserving completely positive maps.
minor comments (4)
  1. [Lemma 5, Eq. (10)] In Eq. (10), the unitary U is defined on AB only, but it is applied to states on RAB; the expression should be written with an explicit extension such as (1_R⊗U)(ρ_RA⊗ω_B)(1_R⊗U†) or a convention should be stated.
  2. [Lemma 8, proof around Eqs. (23)-(24)] The derivation of Eq. (24) is compressed into 'using the same argument'; since Eq. (24) is essential for the application of Lemma 7, adding one or two intermediate lines would make the proof easier to verify.
  3. [Appendix A, Eq. (A10) and surrounding diagrams] The tensor-network proof of Proposition 10 is difficult to parse in printed form; a sentence explaining the diagrammatic conventions, including the role of the suppressed W_j and the half-circles, would improve readability.
  4. [Fig. 1 caption] The caption states 'all inclusions are strict' without qualification; it should point to the conditional status of EQ∩DS⊊F that is discussed in the text and footnote [40].

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the equilibrating-dilation characterization of thermal operations is derived from an entropy argument, with the Gibbs identification applied after the fact; the MIP*=RE-dependent strict inclusion in the hierarchy is explicitly flagged in the paper's own footnote [40] as unpeer-reviewed, an external caveat rather than a circular step.

full rationale

Definition 1 defines an equilibrating dilation purely information-theoretically through Eq. (3), the invariance of the environment marginal. Proposition 4 is derived, not assumed: Lemma 3 uses unitary invariance and additivity of the von Neumann entropy together with nonnegativity of mutual information to conclude I(A:B)=0 and hence U(omega_A tensor omega_B)U^dagger = omega_A tensor omega_B from the marginal condition Eq. (1). Interpreting the full-rank states omega_A and omega_B as Gibbs states via H_A = -beta^{-1}(log omega_A + log Z_A) is an identification performed after the fact, not an input assumption. Combining this with [U, omega_A tensor omega_B]=0 gives [U, H_A+H_B]=0, which is exactly the defining condition of a thermal operation. Conversely, an energy-preserving dilation of a thermal operation trivially satisfies Eq. (3), so the claimed 'precisely' correspondence is a genuine equivalence with the nontrivial direction proved. The self-citations appearing in the paper are not load-bearing in a circular way: Proposition 10 taken from [19] is re-proved in full in Appendix A, and the doubly-stochastic property of catalytic channels taken from [18] is an external peer-reviewed fact whose stated assumptions do not include the present target results. Lemma 8's proof uses monotonicity of quantum relative entropy and not a chain of self-citations. No fitted parameter is relabeled as a prediction, and no quantity is defined in terms of the result it is used to establish. Two caveats are weighed in this verdict. First, the strict inclusion EQ intersect DS subset F (Eq. 15, printed as part of Eq. 12 and Fig. 1) depends on MIP*=RE [15], which footnote [40] itself says has no completely peer-reviewed proof, and the main text hedges with 'If true' immediately before concluding Eq. (15); this is an external-verification and correctness risk rather than a circular step, because the cited theorem is not authored here and no equation reduces to its own input. Second, the thermodynamic identification is stated only for full-rank fixed points, a scope restriction the paper discloses explicitly. Neither caveat constitutes circularity, so the circularity score is 0.

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

No fitted parameters: the inverse temperature β is a chosen convention, not fitted, and no numerically optimized constants appear. No new physical entities are postulated; informational equilibrium and the various dilation classes are mathematical definitions. The heaviest external input is MIP*=RE for one strict inclusion in the hierarchy.

assumptions (6)
  • domain assumption All Hilbert spaces are finite-dimensional.
    Stated in the Introduction: 'Throughout this work, we only consider finite-dimensional Hilbert spaces.' The entropy and support arguments depend on this restriction.
  • standard math Von Neumann entropy is unitarily invariant, additive over tensor products, and zero mutual information implies a product state.
    Used in the proofs of Lemma 3 and Lemma 7.
  • domain assumption Full-rank density matrices can be interpreted as Gibbs states by defining Hamiltonians a posteriori.
    Used in the section 'Equilibration and thermal operations' with H_A := -β^{-1}(log ω_A + log Z_A); this identifies equilibrating channels with thermal operations.
  • domain assumption MIP*=RE, giving a negative resolution of the Connes embedding problem, is valid.
    Used for the strict inclusion EQ∩DS⊊F (Eq. 15); footnote [40] notes no completely peer-reviewed proof exists at the time of writing.
  • standard math Quantum relative entropy is monotone under channels and superadditive for product reference states.
    Used in the proof of Lemma 8.
  • domain assumption Robust catalysis requires the catalyst state to be exactly preserved for all input states.
    Borrowed from [34] and used in Lemma 8; this is a strong but standard definition in the cited work.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Thermal operations from informational equilibrium." pith.science (2026). https://pith.science/paper/V5QP5QID

@misc{pith2026250716637,
  author       = {Pith},
  title        = {Pith review of: Thermal operations from informational equilibrium},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/V5QP5QID}},
  note         = {Machine review of arXiv:2507.16637}
}
read the original abstract

Thermal operations are quantum channels that have taken a prominent role in deriving fundamental thermodynamic limitations in quantum systems. We show that these channels are uniquely characterized by a purely quantum information theoretic property: They admit a dilation into a unitary process that leaves the environment invariant when applied to the equilibrium state. In other words, they are the only channels that preserve equilibrium between system and environment. Extending this perspective, we explore an information theoretic idealization of heat bath behavior, by considering channels where the environment remains locally invariant for every initial state of the system. These are known as catalytic channels. We show that catalytic channels provide a refined hierarchy of Gibbs-preserving maps for fully-degenerate Hamiltonians, and are closely related to dual unitary quantum circuits.

Figures

Figures reproduced from arXiv: 2507.16637 by the authors.

Figure 1
Figure 1. FIG. 1. A hierarchy of doubly-stochastic channels, where all in [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

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

  1. Gaussian time-translation covariant operations: structure, implementation, and thermodynamics

    quant-ph 2026-01 accept novelty 7.0 of 10

    Gaussian time-translation-covariant operations are classified; freely dilatable ones are characterized by AA†≤I and supp(B)=supp(I−AA†), and Gaussian enhanced thermal operations coincide with Gaussian thermal operations.

Reference graph

Works this paper leans on

63 extracted references · 59 canonical work pages · cited by 1 Pith paper

  1. [40]

    Y . Ding, F. Ding, and X. Hu, Phys. Rev. A103, 052214 (2021)

    work page 2021

  2. [15]

    Faist and R

    P. Faist and R. Renner, Physical Review X 8, 021011 (2018)

    work page 2018

  3. [1]

    Mixed unitary channels (MU),

    work page

  4. [2]

    for every t ∈ R and every state ρA on A we have T (ωit AρAω−it A ) = ωit AT (ρA)ω−it A ,

    work page

  5. [3]

    The third statement of Proposition 4 facilitates the thermo- dynamic interpretation of equilibrating channels

    if ωA has full rank, ωB can be chosen to have full rank. The third statement of Proposition 4 facilitates the thermo- dynamic interpretation of equilibrating channels. When a sys- tem is in thermal equilibrium with a heat bath, its equilibrium state is the Gibbs state e−βH /Z corresponding to the system Hamiltonian H and the bath temperature β−1 = kBT . T...

    work page

  6. [4]

    Catalytic channels (CAT),

    work page

  7. [5]

    Doubly-stochastic equilibrating channels (EQ ∩ DS),

    work page

  8. [6]

    We assume that the catalytic and equilibrating dilating sys- tem is finite-dimensional

    Factorizable channels (F). We assume that the catalytic and equilibrating dilating sys- tem is finite-dimensional. These classes of doubly-stochastic quantum channels now form a hierarchy: MU ⊊ CAT ⊊ EQ ∩ DS ⊊ F ⊊ DS, (12) wherein all inclusions are strict. In particular, the strict in- clusions MU ⊊ CAT ⊊ EQ were previously open problems [18, 19] and are...

    work page

Show all 63 references
  1. [7]

    T is a mixed unitary channel

  2. [8]

    T admits a non-degenerate equilibrating dilation

  3. [9]

    Quantum Metrology Theory Based on Temporal Correlation Opera- tors

    T admits a non-degenerate catalytic dilation. Proof. 3 → 2 follows from the definition of the catalytic di- lation. 1 → 3 can be proved using Eq. (6), because we can always split up probabilities into distinct summands and dis- tribute them over more dimensions to make the cat...

    2025

  4. [10]

    Horodecki and J

    M. Horodecki and J. Oppenheim, Nature Communications 4, 2059 (2013)

    2013

  5. [11]

    F. G. S. L. Brand˜ao, M. Horodecki, J. Oppenheim, J. M. Renes, and R. W. Spekkens, Physical Review Letters 111, 250404 (2013)

    2013

  6. [13]

    Faist, F

    P. Faist, F. Dupuis, J. Oppenheim, and R. Renner, Nature Com- munications 6, 7669 (2015)

    2015

  7. [14]

    Faist, J

    P. Faist, J. Oppenheim, and R. Renner, New Journal of Physics 17, 043003 (2015)

    2015

  8. [16]

    Lostaglio, D

    M. Lostaglio, D. Jennings, and T. Rudolph, Nature Communi- cations 6, 6383 (2015)

    2015

  9. [17]

    Tajima and R

    H. Tajima and R. Takagi, Physical Review Letters 134, 170201 (2025)

    2025

  10. [18]

    Janzing, P

    D. Janzing, P. Wocjan, R. Zeier, R. Geiss, and T. Beth, Int. J. Th. Phys. 39, 2717 (2000)

    2000

  11. [19]

    Lostaglio, K

    M. Lostaglio, K. Korzekwa, D. Jennings, and T. Rudolph, Phys. Rev. X 5, 021001 (2015)

    2015

  12. [20]

    ´Cwikli´nski, M

    P. ´Cwikli´nski, M. Studzi ´nski, M. Horodecki, and J. Oppen- heim, Physical Review Letters 115, 210403 (2015)

    2015

  13. [21]

    Faist, J

    P. Faist, J. Oppenheim, and R. Renner, New J. Phys.17, 043003 (2015)

    2015

  14. [22]

    Haagerup and M

    U. Haagerup and M. Musat, Communications in Mathematical Physics 303, 555 (2011)

    2011

  15. [23]

    Haagerup and M

    U. Haagerup and M. Musat, Communications in Mathematical Physics 338, 721 (2015)

    2015

  16. [24]

    MIP*=RE,

    Z. Ji, A. Natarajan, T. Vidick, J. Wright, and H. Yuen, “MIP*=RE,” (2022), arxiv:2001.04383

    2022 arXiv

  17. [25]

    Lipka-Bartosik, H

    P. Lipka-Bartosik, H. Wilming, and N. H. Ng, Reviews of Mod- ern Physics 96, 025005 (2024)

    2024

  18. [26]

    P. Boes, H. Wilming, R. Gallego, and J. Eisert, Physical Review X 8, 041016 (2018)

    2018

  19. [27]

    S. H. Lie and H. Jeong, Physical Review Research 3, 013218 (2021)

    2021

  20. [28]

    S. H. Lie and H. Jeong, Phys. Rev. Res. 3, 043089 (2021)

    2021

  21. [29]

    S. H. Lie and H. Jeong, Phys. Rev. A 107, 042430 (2023). 6

    2023

  22. [30]

    Anantharaman-Delaroche, Probability Theory and Related Fields 135, 520 (2006)

    C. Anantharaman-Delaroche, Probability Theory and Related Fields 135, 520 (2006)

    2006

  23. [31]

    Birkhoff, Univ

    G. Birkhoff, Univ. Nac. Tacuman, Rev. Ser. A5, 147 (1946)

    1946

  24. [32]

    Piroli, B

    L. Piroli, B. Bertini, J. I. Cirac, and T. Prosen, Physical Review B 101, 094304 (2020)

    2020

  25. [33]

    Foligno, P

    A. Foligno, P. Kos, and B. Bertini, Physical Review Letters 132, 250402 (2024)

    2024

  26. [34]

    Borsi and B

    M. Borsi and B. Pozsgay, Physical Review B 106, 014302 (2022)

    2022

  27. [35]

    Keyl and R

    M. Keyl and R. F. Werner, J. Math. Phys. 40, 3283 (1999)

    1999

  28. [36]

    Marvian and R

    I. Marvian and R. W. Spekkens, Nat. Commun. 5, 3821 (2014)

    2014

  29. [37]

    Resource theory of quan- tum thermodynamics: Thermal operations and second laws,

    N. H. Y . Ng and M. P. Woods, “Resource theory of quan- tum thermodynamics: Thermal operations and second laws,” (Springer International Publishing, Cham, 2018) pp. 625–650

    2018

  30. [38]

    Lostaglio, Rep

    M. Lostaglio, Rep. Prog. Phys. 82, 114001 (2019)

    2019

  31. [39]

    Gibbs-preserving operations re- quiring infinite amount of quantum coherence,

    H. Tajima and R. Takagi, “Gibbs-preserving operations re- quiring infinite amount of quantum coherence,” (2024), arxiv:2404.03479

    2024 arXiv

  32. [41]

    F. G. S. L. Brand˜ao, M. Horodecki, N. H. Y . Ng, J. Oppenheim, and S. Wehner, PNAS112, 3275 (2015)

    2015

  33. [42]

    Kretschmann, D

    D. Kretschmann, D. Schlingemann, and R. F. Werner, IEEE Transactions on Information Theory 54, 1708 (2008)

    2008

  34. [43]

    Robust catalysis and resource broadcasting: The possible and the impossible,

    J. Son, R. Ganardi, S. Minagawa, F. Buscemi, S. H. Lie, and N. H. Ng, “Robust catalysis and resource broadcasting: The possible and the impossible,” (2024), arXiv:2412.06900

    2024

  35. [44]

    Wilming, Physical Review Letters 127, 260402 (2020), arXiv:2012.05573 [quant-ph]

    H. Wilming, Physical Review Letters 127, 260402 (2020), arXiv:2012.05573 [quant-ph]

    2020 arXiv

  36. [45]

    Wilming, Quantum 6, 858 (2022)

    H. Wilming, Quantum 6, 858 (2022)

    2022

  37. [46]

    The structure of unital maps and the asymptotic quan- tum birkhoff conjecture,

    P. Shor, “The structure of unital maps and the asymptotic quan- tum birkhoff conjecture,” (2010), Steklov Mathematical Insti- tute Seminar

    2010

  38. [47]

    G. Gour, M. P. M¨uller, V . Narasimhachar, R. W. Spekkens, and N. Yunger Halpern, Phys. Rep. The resource theory of informa- tional nonequilibrium in thermodynamics, 583, 1 (2015)

    2015

  39. [48]

    Connes, Annals of Mathematics 104, 73 (1976)

    A. Connes, Annals of Mathematics 104, 73 (1976)

    1976

  40. [49]

    At the time of writing, no completely peer-reviewed proof of

  41. [50]

    Haagerup, M

    U. Haagerup, M. Musat, and M. B. Ruskai, Annales Henri Poincar´e 22, 3455 (2021), arxiv:2006.03414

    2021 arXiv

  42. [51]

    Chitambar and G

    E. Chitambar and G. Gour, Rev. Mod. Phys.91, 025001 (2019)

    2019

  43. [52]

    Lostaglio, ´A

    M. Lostaglio, ´A. M. Alhambra, and C. Perry, Quantum 2, 52 (2018)

    2018

  44. [53]

    Lostaglio and K

    M. Lostaglio and K. Korzekwa, Phys. Rev. A 106, 012426 (2022)

    2022

  45. [54]

    Son and N

    J. Son and N. H. Ng, New Journal of Physics26, 033029 (2024)

    2024

  46. [55]

    Son and N

    J. Son and N. H. Ng, Quantum Science and Technology 10, 015011 (2024)

    2024

  47. [56]

    Araki and E

    H. Araki and E. H. Lieb, Communications in Mathematical Physics 18, 160 (1970)

    1970

  48. [57]

    Dykema and K

    K. Dykema and K. Juschenko, Mathematica Scandinavica 109, 225 (2011). END MATTER Proofs of main results.— Hereby we present the proofs of Lemma 3 and Proposition 4. Proof. (Lemma 3) The von Neumann entropy H(ω) := −Tr(ω log(ω)) is unitarily invariant, additive over tensor- pro...

    2011

  49. [58]

    U ⊤A is unitary and U (1 ⊗ σB)U † = 1 ⊗ (W †σBW ) for a unitary W on B

  50. [59]

    ((1 ⊗ W )U, σB) is a catalytic dilation

  51. [60]

    In particular, if U ⊤A is unitary, σB yields a catalytic dilation if and only if [U, 1 ⊗ σB] = 0

    (U, σB) is a non-correlating dilation. In particular, if U ⊤A is unitary, σB yields a catalytic dilation if and only if [U, 1 ⊗ σB] = 0. We here provide a diagrammatic proof of Proposition 10 for completeness. Let us first recall a simple observation about the partial transpos...

  52. [61]

    Let T admit an exact factorization that is also a catalytic dilation

    CAT ⊊ EQ Lemma 13. Let T admit an exact factorization that is also a catalytic dilation. Then either T is unitary, or T is a non-trivial convex mixture of doubly-stochastic quantum channels. In other words,T is not extremal in the set of doubly-stochastic channels if it is not...

  53. [62]

    To do so, we first need to introduce the notion of Schur multipliers

    MU ⊊ CAT Next, we detail some technical ingredients used to show the strict inclusion of mixed unitary channels in the set of catalytic dilations. To do so, we first need to introduce the notion of Schur multipliers. In the following, we denote by Mn(C) (Mn(R)) the set of n × ...

  54. [63]

    The Schur multiplier TX is exactly factorizable as TX (ρ) = Tr2(U ρ⊗ 1 2d U †), where U = P j |j ⟩⟨j | ⊗uj

  55. [64]

    The pair (U, 1/2d) is a catalytic dilation of TX. Proof. The proof combines several observations in [13]. First, by [13, Remark 2.7] we can write TB(x) = Pd i=1 aixai, where ai ∈ Mn(R) are real, diagonal matrices that are linearly independent and fulfill Pd i=1 a2 i = 1. We no...

Pith tools

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