Pith. sign in

REVIEW 3 major objections 5 minor 67 references

The paper claims that causal order in process-matrix quantum mechanics is subsystem-dependent: a global redefinition of agents can turn many acausal processes into causally ordered ones, and the spectrum alone decides exactly when this is p

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-01 23:38 UTC pith:Y4CBOWJC

load-bearing objection Solid spectral core, but the 'generic processes' headline rests on Wishart spectra rather than valid process matrices. the 3 major comments →

arxiv 2607.15345 v1 pith:Y4CBOWJC submitted 2026-07-16 quant-ph

Causality from the spectrum: Emergence of causal order from process-matrix mereology

Varun Kushwaha , Nicolas Loizeau , Alexei Grinbaum , Oliver Friedrich This is my paper
classification quant-ph MSC 81P4515A18 PACS 03.67.-a03.65.-w
keywords process matrixcausal orderquantum mereologyspectral characterizationeigenvalue multiplicitiesrandom matrix typicalityemergence of causalitytensor-product decomposition
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper asks whether 'which agent comes before which' can be a derived property rather than a postulate. It works in the process-matrix framework, where a process is a positive operator on a global Hilbert space and agents are defined by a choice of input/output subsystems. It establishes an exact spectral test: a process can be redrawn by a global unitary redefinition of agents as a process with a fixed total causal order if and only if the output dimension of the last agent divides every eigenvalue multiplicity, provided a mild divisibility condition on subsystem dimensions. It then shows that random high-dimensional spectra nearly satisfy this test: the expected squared distance to the total-order orbit is at most 2D/(D^2+1), which vanishes as the Hilbert-space dimension grows. The sympathetic reader takes away a concrete mechanism—spectral flatness, not detailed dynamics—by which classical causal order could emerge from algebraic data alone.

Core claim

The central discovery is that causal order in process-matrix quantum mechanics is not intrinsic to a process but depends on the chosen tensor-product decomposition. Given any positive semidefinite, trace-D_O operator W on the global Hilbert space, a global unitary U redefines the agents and their input/output spaces while preserving the spectrum. Theorem 1 states that, under the chain-divisibility condition d_{k,O} | d_{k+1,I}, W lies in the global-unitary orbit of a valid process with total order 1≺...≺n if and only if d_{n,O} divides every eigenvalue multiplicity of W. Consequently a simple-spectrum process is incompatible with any nontrivial total causal order. In the thermodynamic limit,

What carries the argument

The central object is the eigenvalue-multiplicity divisibility condition of Theorem 1, which is enforced by the 'open final wire' factorization X = 1_{n,O}/d_{n,O} ⊗ K of any fixed-total-order process. Because the last agent has no future, its output Hilbert space acts as an identity factor, forcing every eigenvalue of X to appear d_{n,O} times. The sufficiency direction is carried by a generalized Bell-basis construction along the causal chain: under chain divisibility d_{k,O} | d_{k+1,I}, the previous output is paired with a block of the next input, and bridge states with maximally mixed margins automatically satisfy the recursive fixed-order constraints. Theorem 2 then shows that the Hilb

Load-bearing premise

The conclusion that generic quantum processes become causally orderable assumes that typical valid process matrices have spectra as flat as random Wishart spectra; the paper checks this only for random density matrices, not for matrices drawn from the valid-process constraint set.

What would settle it

Sample valid process matrices uniformly from the convex set defined by positivity, trace-D_O, and the Pauli-support constraints for n qubits with n = 2,...,8, compute the mean of δ^2_π(ρ), and check whether it decays roughly as 2^{1-n}. If the empirical mean stays above a fixed positive ε as n grows, the typicality result does not extend to actual process matrices.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Exact spectral test: any process whose eigenvalue multiplicities are not all divisible by the last agent's output dimension cannot be made totally causally ordered by any global redefinition of agents; a simple spectrum can never be totally ordered.
  • The number of agents is itself part of the freedom: if eigenvalue multiplicities have gcd g(W), the last agent can contain at most ℓ_d(W) elementary output factors, so splitting the last agent can make a spectrum orderable that was not orderable with a merged last agent.
  • Restricted redefinitions behave differently: under unitaries that preserve the input/output split, a process is orderable only if its output operator system Sout(W) can be conjugated into one party's output algebra; a commuting subspace of dimension greater than max(d_{AO}-1, d_{BO}-1) blocks both strict orders.
  • Random high-dimensional spectra almost surely fail the exact degeneracy condition at every finite dimension, yet their mean squared distance to the total-order orbit is at most 2D/(D^2+1), so the distance converges to zero in probability as D grows.
  • The operative mechanism is spectral flatness, not chaos: if purity tr(ρ^2) tends to zero along a dimension-growing family compatible with the divisibility conditions, the distance to the orbit vanishes, whereas spiked high-purity spectra remain at a fixed distance that does not decay with dimension.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • An extension the paper suggests but does not perform: test the typicality claim directly by sampling valid process matrices from a natural measure on the convex set defined by positivity, trace-D_O, and the Pauli-support constraints, then measuring the average spectral distance to the total-order orbit as dimension grows.
  • The result covers only total orders with a unique last agent; a natural extension is to partial causal orders with several 'last' agents, where a generalized divisibility condition on the joint future algebra may replace the d_{n,O} divisor condition.
  • Theorem 2 identifies the closest block-degenerate spectrum but not an efficient construction of the unitary U; this suggests an algorithmic follow-up using gradient or randomized methods on the unitary orbit with the spectral cost function.
  • If spectral flatness is the deciding factor, then indefinite causal order should persist in small or high-purity systems where spectra have spikes, and should become increasingly redefinable as the effective dimension grows.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper studies whether a bipartite or multipartite process matrix W can be made causally ordered by a global unitary redefinition of the tensor-product decomposition — i.e., by redefining the agents and their input/output subsystems. The main results are: (i) an exact spectral criterion (Theorem 1, Eq. 18) under the dimension condition (17), stating that W lies in the global-unitary orbit of a fixed total-order process iff the output dimension of the last agent divides every eigenvalue multiplicity; (ii) an explicit Hilbert–Schmidt distance formula (Theorem 2, Eq. 56) from a normalized spectrum to the normalized total-order orbit, equal to the within-block spectral variance; (iii) numerical and analytic evidence that spectra drawn from the balanced Hilbert–Schmidt induced (Wishart) ensemble approach the total-order orbit in large dimension, leading to the paper's headline claim that generic high-dimensional quantum processes admit a preferred causally ordered decomposition. The paper also derives a same-time unitary obstruction (Proposition 2) and discusses splitting and merging agents in terms of the gcd of eigenvalue multiplicities.

Significance. If the exact results are taken on their own, the paper makes a solid and useful contribution. Theorem 1 gives a clean spectral necessary-and-sufficient condition for compatibility with a total causal order, and the sufficiency construction in Appendix A.2 is non-trivial: it shows that the required spectral degeneracy is not just a necessary obstruction but can always be realized by a valid fixed-order process under the stated dimension assumptions. Theorem 2 provides a simple, operational distance formula that quantifies how far a spectrum is from a total-order orbit, and the paper honestly exhibits a spiked family (Eqs. 59–62) for which the distance does not vanish even as D grows. The code availability and the careful distinction between exact and approximate orderability are also strengths. However, the paper's broader claim that 'generic quantum processes' become causally ordered in the thermodynamic limit rests on an unproven identification of typical process-matrix spectra with Wishart spectra. This is the main load-bearing weakness.

major comments (3)
  1. [Sec. III.E and III.G; Eq. (49) and Eq. (65)] The step from Wishart spectra to 'generic quantum processes' is unsupported. The sampled operator ρ = GG†/tr(GG†) is, as the paper itself states in Sec. III.E, 'used here only to supply a random spectrum' and 'need not itself satisfy the process-matrix constraints.' The bound E[δ^2] ≤ 2D/(D^2+1) in Eq. (65) therefore concerns the balanced induced ensemble of density matrices, not any measure over valid process matrices. This is not a harmless distinction: using the construction in Appendix A.2, for a fixed total order with final output dimension d_f, one can choose K = D_O |e_J⟩⟨e_J| for a bridge-basis element and form the valid ordered process W = (1_{n,O}/d_f) ⊗ K. The normalized process ρ = W/D_O has purity 1/d_f independent of the total dimension D. Thus high-dimensional valid process matrices can be strongly non-flat. Without a specified probability measure over valid process matric
  2. [Sec. IV, final paragraph; Sec. III.G convergence claim (Eqs. 67–68)] The paper's headline conclusion — 'typical high-dimensional process matrices have flat random spectra' and therefore 'generic quantum processes admit a preferred decomposition with a definite causal order' — does not follow from the Wishart analysis. The convergence in probability in Eq. (68) is a statement about random density matrices drawn from the balanced induced ensemble. The set of valid process matrices is a lower-dimensional convex subset satisfying additional linear constraints, and the paper offers no argument that any natural measure on this set produces spectra with the same low-purity behavior. The spiked family in Eqs. (59–62) shows that low-dimensional spectral concentration can persist at arbitrarily large D, so dimension alone is not sufficient. To support the generic claim, the authors would need either to exhibit a measure on actual process matrices and prove a concen
  3. [Sec. II, Eq. (5) and Sec. III.E, numerical cost] The numerical minimization uses the spectral cost C(w) of Eq. (4) and the Hoffman–Wielandt bound in Eq. (6) to bound the forbidden weight of the recovered operator. This is methodologically reasonable, but it should be stated more carefully that the numerical procedure minimizes over operators supported on the allowed Pauli set, and the recovery U = V′V† is only guaranteed to make U W U† close to the fitted operator when the residual is small; Eq. (6) controls the error but does not by itself certify that the final state is a valid process matrix. The paper does acknowledge this after Eq. (6), so this is a presentation concern rather than a technical error.
minor comments (5)
  1. [Fig. 3 caption] The y-axis label appears garbled: 'n |En n|^2' should presumably be Σ_n (E_n − E_n(w))² or the corresponding spectral cost. Please correct the typo.
  2. [Sec. III.E, first sentence] The sentence 'The Wishart draw is used here only to supply a random spectrum' is important for honesty, but given the major concern above, it should be moved and expanded so that the reader understands that the numerical computation is not sampling valid process matrices.
  3. [Eq. (2)] The expression for W_GYNI has a small formatting issue: the fraction inside the parentheses is ambiguous. Please add explicit brackets.
  4. [References] Reference [28] (Brukner, Nature Physics) and reference [26] lack complete publication data; please add volume/page or DOI for consistency with the other references.
  5. [General notation] The notation Oπ and bOπ is clear, but it would help to state explicitly that bOπ consists of normalized operators X/D_O and that the infimum in Eq. (51) is over normalized fixed-order processes, since the trace normalization differs between W and ρ.

Circularity Check

0 steps flagged

No circular step in the spectral derivation; only minor non-load-bearing self-citations, while the Wishart-to-'generic process' inference is an unproven typicality premise rather than a circular reduction.

full rationale

The paper's derivation chain is not circular in the sense of the seven patterns. Theorem 1 (Eq. 18) is derived from the fixed-order factorization X = 1_{n,O}/d_{n,O} ⊗ K (Eq. 16) plus an explicit Bell-bridge eigenbasis construction in Appendix A.2; sufficiency is constructive, not a restatement of the criterion. Theorem 2 (Eq. 56) follows from Theorem 1 by a spectral rearrangement/Hoffman-Wielandt argument (Appendix A.4). The OCB and Lugano examples are checked by explicit spectra. The only self-references are to the authors' earlier mereology optimization methods (refs [8]-[10]); they are used for numerical algorithms and are not load-bearing for the spectral theorems. The genuinely fragile step is the unproven identification of typical valid-process spectra with Wishart spectra: Sec. III.E explicitly states the Wishart draw 'need not itself satisfy the process-matrix constraints,' and Sec. IV asserts without proof that 'typical high-dimensional process matrices have flat random spectra.' That is an unsupported typicality premise - a correctness/validity gap, not a fitted-parameter prediction or definitional equivalence. Hence no circular step is identified; the score of 2 reflects only minor non-load-bearing self-citations.

Axiom & Free-Parameter Ledger

0 free parameters · 5 axioms · 0 invented entities

The paper's rigorous core adds one assumption beyond standard process-matrix theory: the chain-divisibility condition (17). The large-dimension claim additionally assumes Wishart spectra are representative of valid process spectra — an unproven modeling choice. No free parameters are fitted and no new physical entities are introduced.

axioms (5)
  • domain assumption Valid process matrices are exactly operators W≥0 with tr W = D_O supported on the allowed Pauli set P (or C for a fixed order).
    Sec. II and Sec. I.B; this is the Oreshkov-Costa-Brukner characterization [17] used to reduce the problem to spectral support on C.
  • domain assumption Fixed-order process constraints are the recursive replacement constraints [1−O_k]F_k K=0 with X=1_{n,O}/d_{n,O}⊗K.
    Eq. (16), Appendix A1; standard fixed-order characterization from [17,37,43].
  • domain assumption Chain-divisibility condition d_{k,O} | d_{k+1,I} for k=1..n−1 is required for sufficiency of the spectral criterion.
    Eq. (17); required for the generalized Bell-basis construction in Appendix A2; without it Theorem 1 gives only necessity.
  • ad hoc to paper The balanced Hilbert-Schmidt induced (Wishart) ensemble represents typical process-matrix spectra.
    Sec. III.E samples rho=GG^dagger/tr(GG^dagger) and explicitly says it need not satisfy process constraints; the paper later states typical process matrices are flat without proof.
  • standard math Spectra of absolutely continuous random-matrix ensembles are simple almost surely.
    Used in Sec. III.G to state Wishart spectra fail exact degeneracy at finite D; cited to [46,47].

pith-pipeline@v1.3.0-alltime-deepseek · 22059 in / 18555 out tokens · 193005 ms · 2026-08-01T23:38:00.814817+00:00 · methodology

0 comments
read the original abstract

In Hamiltonian systems, the only basis-independent quantity is the spectrum. Given a spectrum, quantum mereology seeks preferred tensor-product decompositions into local subsystems, leading to the emergence of locality. Causal order, however, remains fixed by the Hamiltonian time evolution. In contrast, higher-order quantum theory permits more general processes that need not possess a definite global causal order. In the process-matrix framework, specifying a process requires a choice of subsystems corresponding to the input and output Hilbert spaces of each agent. A change of basis redefines both the agents and the corresponding decompositions into subsystems, while leaving the spectrum of the process matrix invariant. Here, we study how causality arises from this spectrum. First, we derive spectral constraints on processes compatible with definite causal order. Second, we show that, in the thermodynamic limit, generic quantum processes admit a preferred decomposition with a definite causal order. This suggests a mechanism for the emergence of classical causality only from algebraic ingredients.

Figures

Figures reproduced from arXiv: 2607.15345 by Alexei Grinbaum, Nicolas Loizeau, Oliver Friedrich, Varun Kushwaha.

Figure 1
Figure 1. Figure 1: FIG. 1 [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2 [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3. Minimum of the spectral cost in Eq. ( [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4. Representation of the optimization routes the dif [PITH_FULL_IMAGE:figures/full_fig_p019_4.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

67 extracted references · 4 canonical work pages

  1. [1]

    S. M. Carroll and A. Singh, Mad-dog everettianism: Quantum mechanics at its most minimal, inWhat is Fun- damental?, edited by A. Aguirre, B. Foster, and Z. Mer- ali (Springer International Publishing, Cham, 2019) pp. 95–104

    2019

  2. [2]

    S. M. Carroll and A. Singh, Phys. Rev. A103, 022213 (2021)

    2021

  3. [3]

    Sels and M

    D. Sels and M. Wouters, Quantum equivalence, the sec- ond law and emergent gravity (2014), arXiv:1411.3901 [quant-ph]

    Pith/arXiv arXiv 2014

  4. [4]

    Tegmark, Chaos, Solitons & Fractals76, 238 (2015)

    M. Tegmark, Chaos, Solitons & Fractals76, 238 (2015)

    2015

  5. [5]

    Grinbaum, British Journal for the Philosophy of Sci- ence58, 387 (2007)

    A. Grinbaum, British Journal for the Philosophy of Sci- ence58, 387 (2007)

    2007

  6. [6]

    Bohr,Atomic Theory and the Description of Nature (Cambridge University Press, 1934)

    N. Bohr,Atomic Theory and the Description of Nature (Cambridge University Press, 1934)

    1934

  7. [7]

    Taranto, S

    P. Taranto, S. Milz, M. Murao, M. T. Quintino, and K. Modi, Higher-order quantum operations (2025), arXiv:2503.09693 [quant-ph]

    Pith/arXiv arXiv 2025

  8. [8]

    Loizeau, F

    N. Loizeau, F. Morone, and D. Sels, Proceedings of the National Academy of Sciences120, e2308006120 (2023), https://www.pnas.org/doi/pdf/10.1073/pnas.2308006120

    work page doi:10.1073/pnas.2308006120 2023

  9. [9]

    C. Cao, O. Friedrich, M. Girard, N. Loizeau, and A. Singh, Wave packets from the spectrum (2026), arXiv:2606.10004 [quant-ph]

    Pith/arXiv arXiv 2026

  10. [10]

    Loizeau and D

    N. Loizeau and D. Sels, Foundations of Physics55, 3 (2024)

    2024

  11. [11]

    Zanardi, E

    P. Zanardi, E. Dallas, F. Andreadakis, and S. Lloyd, Quantum8, 1406 (2024)

    2024

  12. [12]

    J. S. Cotler, G. R. Penington, and D. H. Ranard, Com- munications in Mathematical Physics368, 1267 (2019)

    2019

  13. [13]

    Soulas, G

    A. Soulas, G. Franzmann, and A. di Biagio, On the emer- gence of preferred structures in quantum theory (2025), arXiv:2512.07468 [quant-ph]

    arXiv 2025

  14. [14]

    Grinbaum, Foundations of Physics45, 1341 (2015)

    A. Grinbaum, Foundations of Physics45, 1341 (2015)

    2015

  15. [15]

    Grinbaum, inNew Challenges to Philosophy of Sci- ence, The Philosophy of Science in a European Perspec- tive, Vol

    A. Grinbaum, inNew Challenges to Philosophy of Sci- ence, The Philosophy of Science in a European Perspec- tive, Vol. 4, edited by H. Andersen, D. Dieks, W. J. Gon- zalez, T. Uebel, and G. Wheeler (Springer, Amsterdam,

  16. [16]

    S. M. Carroll, Reality as a vector in hilbert space, in Quantum Mechanics and Fundamentality : Naturalizing Quantum Theory between Scientific Realism and Onto- logical Indeterminacy, edited by V. Allori (Springer In- ternational Publishing, Cham, 2022) pp. 211–224

    2022

  17. [17]

    Oreshkov, F

    O. Oreshkov, F. Costa, and ˇC. Brukner, Nature Commu- nications3, 1092 (2012)

    2012

  18. [18]

    Wechs, H

    J. Wechs, H. Dourdent, A. A. Abbott, and C. Branciard, PRX Quantum2, 030335 (2021)

    2021

  19. [19]

    Araujo, C

    M. Araujo, C. Branciard, F. Costa, A. Feix, C. Gi- armatzi, and v. Brukner, New Journal of Physics17, 102001 (2015)

    2015

  20. [20]

    Branciard, M

    C. Branciard, M. Araujo, A. Feix, F. Costa, and ˇC. Brukner, New Journal of Physics18, 013008 (2015)

    2015

  21. [21]

    Abbott, C

    A. Abbott, C. Giarmatzi, F. Costa, and C. Branciard, Phys. Rev. A94, 032131 (2016), arXiv:1608.01528

    Pith/arXiv arXiv 2016

  22. [22]

    M. L. Almeida, J.-D. Bancal, N. Brunner, A. Ac ´ ın, N. Gisin, and S. Pironio, Phys. Rev. Lett.104, 230404 (2010)

    2010

  23. [23]

    Baumeler, A

    A. Baumeler, A. Feix, and S. Wolf, Phys. Rev. A90, 042106 (2014)

    2014

  24. [24]

    Baumeler and S

    A. Baumeler and S. Wolf, New Journal of Physics18, 013036 (2016)

    2016

  25. [25]

    van der Lugt, J

    T. van der Lugt, J. Barrett, and G. Chiribella, Device- independent certification of indefinite causal order in the quantum switch (2023), arXiv:2208.00719

    Pith/arXiv arXiv 2023

  26. [26]

    Costa, G

    F. Costa, G. Rubino, C. Branciard, ˇCaslav Brukner, and M. T. Quintino, Indefinite quantum causality (2026), arXiv:2606.19438 [quant-ph]

    Pith/arXiv arXiv 2026

  27. [27]

    Rovelli, Int

    C. Rovelli, Int. J. of Theor. Phys.35, 1637 (1996)

    1996

  28. [28]

    Brukner, Nature Physics 10.1038/s41567-020-0984-8 (2020)

    ˇC. Brukner, Nature Physics 10.1038/s41567-020-0984-8 (2020)

    work page doi:10.1038/s41567-020-0984-8 2020

  29. [29]

    Vilasini and R

    V. Vilasini and R. Renner, Physical Review Letters133 (2024)

    2024

  30. [30]

    Salzger and V

    M. Salzger and V. Vilasini, New Journal of Physics27, 13 023002 (2025), arXiv:2404.05319

    Pith/arXiv arXiv 2025

  31. [31]

    P. A. Gu´ erin and ˇC. Brukner, New Journal of Physics 20, 103031 (2018), arXiv:1805.122429

    Pith/arXiv arXiv 2018

  32. [32]

    Kabel, A.-C

    V. Kabel, A.-C. de la Hamette, L. Apadula, C. Cepollaro, H. Gomes, J. Butterfield, and ˇC. Brukner, Communica- tions Physics8, 10.1038/s42005-025-02084-3 (2025)

    doi:10.1038/s42005-025-02084-3 2025

  33. [33]

    Apadula, A

    L. Apadula, A. Grinbaum, and ˇC. Brukner, Frame perspectives for process matrices: from coordinate parametrization to spacetime representation (2026), arXiv:2604.02873 [quant-ph]

    Pith/arXiv arXiv 2026

  34. [34]

    Grinbaum, Agency under indefinite causalitengel- hardt:2014gcay: operational eternalism in higher-order quantum theory (2025), arXiv:2512.22879 [quant-ph]

    A. Grinbaum, Agency under indefinite causalitengel- hardt:2014gcay: operational eternalism in higher-order quantum theory (2025), arXiv:2512.22879 [quant-ph]

    Pith/arXiv arXiv 2025

  35. [35]

    Oreshkov, Quantum3, 206 (2019), 1801.07594

    O. Oreshkov, Quantum3, 206 (2019), 1801.07594

    Pith/arXiv arXiv 2019

  36. [36]

    Wechs and O

    J. Wechs and O. Oreshkov, Subsystem decompositions of quantum evolutions and transformations between causal perspectives (2024), arXiv:2411.16504 [quant-ph]

    Pith/arXiv arXiv 2024

  37. [38]

    Ara´ ujo, A

    M. Ara´ ujo, A. Feix, M. Navascu´ es, and ˇC. Brukner, Quantum1, 10 (2017)

    2017

  38. [39]

    Costa and S

    F. Costa and S. Shrapnel, New Journal of Physics18, 063032 (2016)

    2016

  39. [40]

    Oreshkov and C

    O. Oreshkov and C. Giarmatzi, New Journal of Physics 18, 093020 (2016)

    2016

  40. [41]

    Wechs, A

    J. Wechs, A. A. Abbott, and C. Branciard, New Journal of Physics21, 013027 (2019)

    2019

  41. [42]

    Vilasini and R

    V. Vilasini and R. Renner, Physical Review A110, 022227 (2024)

    2024

  42. [43]

    Branciard, Scientific Reports6, 10.1038/srep26018 (2016)

    C. Branciard, Scientific Reports6, 10.1038/srep26018 (2016)

    work page doi:10.1038/srep26018 2016

  43. [44]

    Kunjwal and ¨A

    R. Kunjwal and ¨A. Baumeler, Physical Review Letters 131, 10.1103/physrevlett.131.120201 (2023)

    work page doi:10.1103/physrevlett.131.120201 2023

  44. [45]

    Steffinlongo and H

    A. Steffinlongo and H. Dourdent, Phys. Rev. Res.8, 013127 (2026)

    2026

  45. [46]

    M. L. Mehta,Random Matrices, 3rd ed., Pure and Ap- plied Mathematics, Vol. 142 (Elsevier Academic Press, 2004)

    2004

  46. [47]

    P. J. Forrester,Log-Gases and Random Matrices, Lon- don Mathematical Society Monographs Series, Vol. 34 (Princeton University Press, 2010)

    2010

  47. [48]

    Zyczkowski and H.-J

    K. Zyczkowski and H.-J. Sommers, Journal of Physics A: Mathematical and General34, 7111–7125 (2001)

    2001

  48. [49]

    D’Alessio, Y

    L. D’Alessio, Y. Kafri, A. Polkovnikov, and M. Rigol, Advances in Physics65, 239 (2016), https://doi.org/10.1080/00018732.2016.1198134

    arXiv 2016

  49. [50]

    J. M. Maldacena, Adv. Theor. Math. Phys.2, 231 (1998), also published in Int. J. Theor. Phys. 38 (1999) 1113– 1133, arXiv:hep-th/9711200 [hep-th]

    Pith/arXiv arXiv 1998

  50. [51]

    Ryu and T

    S. Ryu and T. Takayanagi, Phys. Rev. Lett.96, 181602 (2006), arXiv:hep-th/0603001 [hep-th]

    Pith/arXiv arXiv 2006

  51. [52]

    Engelhardt and A

    N. Engelhardt and A. C. Wall, Journal of High Energy Physics2015, 73 (2015)

    2015

  52. [53]

    B. S. DeWitt, Phys. Rev.160, 1113 (1967)

    1967

  53. [54]

    J. A. Wheeler, Battelle Rencontres (1968)

    1968

  54. [55]

    C. J. Isham, NATO Sci. Ser. C409, 157 (1993), arXiv:gr- qc/9210011 [gr-qc]

    arXiv 1993

  55. [56]

    Cao and S

    C. Cao and S. M. Carroll, Phys. Rev. D97, 086003 (2018), arXiv:1712.02803 [hep-th]

    Pith/arXiv arXiv 2018

  56. [57]

    Mohan and L

    V. Mohan and L. Thorlacius, Spacetime from operator algebras (2026), arXiv:2606.10924 [hep-th]

    Pith/arXiv arXiv 2026

  57. [58]

    Jacobson, Phys

    T. Jacobson, Phys. Rev. Lett.75, 1260 (1995), arXiv:gr- qc/9504004 [gr-qc]

    arXiv 1995

  58. [59]

    Loizeau, J

    N. Loizeau, J. C. Peacock, and D. Sels, SciPost Phys. Codebases , 54 (2025)

    2025

  59. [60]

    Knill, Non-binary unitary error bases and quantum codes (1996), arXiv:quant-ph/9608048

    E. Knill, Non-binary unitary error bases and quantum codes (1996), arXiv:quant-ph/9608048

    Pith/arXiv arXiv 1996

  60. [61]

    Watrous,The Theory of Quantum Information(Cam- bridge University Press, 2018)

    J. Watrous,The Theory of Quantum Information(Cam- bridge University Press, 2018)

    2018

  61. [62]

    A. J. Hoffman and H. W. Wielandt, Duke Mathematical Journal20, 37 (1953)

    1953

  62. [63]

    Bhatia,Matrix Analysis, Graduate Texts in Mathe- matics, Vol

    R. Bhatia,Matrix Analysis, Graduate Texts in Mathe- matics, Vol. 169 (Springer, 1997). Appendix A: Proofs of the spectral results

    1997

  63. [64]

    These maps give the standard linear constraints defining process matrices and fixed- order process classes [17, 37, 43]

    Replacement maps and fixed-order factorization For a subsystemX, define X M := 1X dX ⊗tr X M , [1−X] M :=M− X M .(A1) When several systems appear as a subscript, the replace- ment acts on all of them. These maps give the standard linear constraints defining process matrices and fixed- order process classes [17, 37, 43]. For the total order 1≺2≺ · · · ≺n, ...

  64. [65]

    The idea is easier to see before writing the general case, so we first give the bipartite qubit construction

    Sufficiency of the spectral criterion We prove the sufficiency direction of Theorem 1. The idea is easier to see before writing the general case, so we first give the bipartite qubit construction. The general proof is then the same construction repeated along the ordered chain. 14 Warm-up: two qubit laboratories Consider two qubit laboratories in the orde...

  65. [66]

    Agent redefinition by splitting and merging Let the global Hilbert space be built fromNelemen- tary input-output pairs labelled by 1, . . . , N. A decom- position intomagents is specified by a partition {1, . . . , N}=G1 ⊔G 2 ⊔ · · · ⊔Gm (A52) together with a total order eA1 ≺ eA2 ≺ · · · ≺eAm .(A53) The agent eAa is defined by the setG a, with H eAa,I :=...

  66. [67]

    By Theorem 1, a normalized operator lies in bOπ only if its eigenvalues are constant in consecutive blocks of lengthd π(n),O

    Distance to the fixed-order orbit Letρhave eigenvalues ν1 ≥ · · · ≥νD ≥0, DX i=1 νi = 1.(A69) For the total orderπ, writeN π = D dπ(n),O . By Theorem 1, a normalized operator lies in bOπ only if its eigenvalues are constant in consecutive blocks of lengthd π(n),O. Thus a spectrum in the fixed-order orbit has the form x1, . . . , x1| {z } dπ(n),O times , x...

  67. [68]

    A same-time unitary may redefine the agents inside the input slice and inside the output slice, but it cannot mix inputs with outputs

    Obstruction for same-time unitaries We prove Proposition 2. A same-time unitary may redefine the agents inside the input slice and inside the output slice, but it cannot mix inputs with outputs. The obstruction below comes from this restriction. Let HI :=H AI ⊗ HBI ,H O :=H AO ⊗ HBO ,(A80) and let Ust =U I ⊗U O , U I ∈U(H I ), U O ∈U(H O).(A81) ThusU I ac...