Pith. sign in

REVIEW 4 minor 35 references

No Rényi divergence other than ordinary relative entropy can obey a universal off-diagonal quantum focusing inequality.

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 →

T0 review · grok-4.5

2026-07-10 17:53 UTC pith:YCP5PFGX

load-bearing objection Clean axiomatic no-go: off-diagonal focusing fails for every non-affine Rényi divergence that satisfies the usual QI axioms, so only relative entropy survives as a universal focusing quantity.

arxiv 2607.07799 v1 pith:YCP5PFGX submitted 2026-07-08 hep-th gr-qcquant-ph

No off-diagonal quantum focusing for R\'enyi divergences

classification hep-th gr-qcquant-ph
keywords quantum focusing conjectureRényi divergenceoff-diagonal QNECdata processing inequalitynull quantizationclassical-quantum conditioningrelative entropy
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 quantum focusing conjecture encodes the idea that gravity remains attractive once quantum effects are included. Its off-diagonal part says that the mixed null-shape variation of relative entropy between two distinct null generators never becomes positive. After a diagonal Rényi version of the quantum null energy condition was proved, it became natural to ask whether a full Rényi focusing statement could hold as well. This paper answers that question in the negative for an entire class of Rényi-type divergences: those that are faithful, obey data processing, add under tensor products, and condition blockwise on matched classical–quantum registers. In a free-field null-quantization setup the authors construct regulated states whose off-diagonal finite differences are strictly negative whenever the Rényi parameter is nonzero. Consequently only the ordinary relative-entropy limit can serve as the entropic ingredient of a universal focusing principle.

Core claim

Within the class of Rényi divergences that satisfy data processing, tensor additivity and matched classical–quantum conditioning, every nonzero-parameter member admits states of free null-quantized fields for which the off-diagonal finite difference of the divergence is strictly negative. Hence no universal off-diagonal focusing inequality can hold for those divergences, and only the affine (relative-entropy) limit survives.

What carries the argument

The off-diagonal finite difference Δ₁₂D_γ together with the matched classical–quantum conditioning rule that reduces it to a classical covariance of binary activation patterns on spectator pencils; the sign of that covariance is then controlled by an exclusive or positively correlated block weight, yielding Δ₁₂D_γ < 0 for every γ ≠ 0.

Load-bearing premise

There must exist at least one single-pencil excitation whose Rényi divergence is finite and takes two different values at two different null cuts.

What would settle it

Exhibit a free-field one-pencil state for which every finite non-rigid seed fails (so that D_γ is either infinite or constant under all null deformations) while a nonzero-γ Rényi divergence still satisfies the other axioms; or produce an explicit counter-example to the classical covariance formula that drives the sign of Δ₁₂.

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

If this is right

  • Only ordinary relative entropy (the affine limit) can enter a universal off-diagonal focusing or quantum-null-energy statement built from the listed axioms.
  • Any claimed Rényi quantum focusing conjecture is immediately ruled out once gravity is switched off, so it cannot hold in the full semiclassical theory either.
  • The diagonal Rényi quantum null energy condition remains open for gravitational back-reaction, because the counter-example construction does not obstruct purely diagonal second variations.
  • Among standard Petz, sandwiched and α-z families, Umegaki relative entropy is uniquely selected as the candidate for quantum expansion.

Where Pith is reading between the lines

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

  • The same classical-covariance obstruction will likely exclude other non-affine divergences that satisfy only a weaker form of matched conditioning, suggesting a sharp operational characterisation of relative entropy among all DPI divergences.
  • Because the counter-example lives already in free null-quantized fields, any future positive focusing statement for a new divergence must either abandon tensor additivity or restrict attention to states without classical correlations across pencils.
  • The survival of the Belavkin–Staszewski relative entropy (which is affine under the same conditioning) points to a possible next candidate for null-energy inequalities that the paper deliberately leaves open.

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

0 major / 4 minor

Summary. The paper proves a no-go theorem for off-diagonal quantum focusing of Rényi-type divergences. After recalling the quantum focusing conjecture and its off-diagonal part (which for relative entropy follows from strong subadditivity), the authors ask whether a Rényi generalization can hold. They consider any one-parameter family D_γ of divergences that is faithful, obeys the data-processing inequality, is additive under tensor products, and satisfies the matched classical–quantum conditioning rule (log-sum-exp for γ ≠ 0, ordinary average for the affine limit γ = 0). In a free-field null-quantization setup with finitely many excited pencils, they construct states by correlating two active pencils with spectator vacuum blocks. For any finite non-rigid one-pencil seed (two cuts with finite unequal D_γ values), exclusive activation (γ > 0) or positive correlation (γ < 0) produces a strictly negative off-diagonal finite difference Δ₁₂ D_γ. Thus no universal off-diagonal focusing inequality can hold for any nonzero γ in the class; only the affine (relative-entropy) limit remains viable. The same obstruction covers the max-relative-entropy endpoint. Diagonal Rényi QNEC is left open.

Significance. If correct, the result cleanly isolates Umegaki relative entropy (and its affine relatives such as Belavkin–Staszewski) as the only members of the standard Rényi class that can enter a universal off-diagonal focusing statement. Combined with existing diagonal QNEC/RQNEC proofs and the SSA argument for ordinary relative entropy, it supplies an operational characterization of the entropic ingredient of quantum focusing. The argument is elementary once the axioms are granted, fully explicit, and free of free parameters or circular constructions; the seed existence is reduced to standard DPI-plus-locality reasoning (with a concrete chiral-current HSMI example). This is a sharp, falsifiable no-go that usefully constrains future Rényi generalizations of QFC/QNEC.

minor comments (4)
  1. In the paragraph preceding Theorem 1 the seed-existence argument is clear, but a one-sentence reminder that the same local unitary can be realized simultaneously on two independent pencils (so that the two seeds are identical) would remove any residual ambiguity about the product construction.
  2. Eqs. (34)–(36) and (41)–(44) are elementary once the activation patterns are chosen; a short remark that the same sign can be read off the classical covariance term (29) would make the geometric origin of the obstruction more transparent.
  3. The End Matter derivation of the matched cq rule for α-z Rényi is helpful; a parenthetical note that the same calculation recovers the known Petz and sandwiched cases would aid readers who work only with those families.
  4. Typographical: “focussing” appears once in the Acknowledgements; standardize to “focusing”.

Circularity Check

0 steps flagged

No significant circularity: the no-go is an elementary finite-difference obstruction constructed directly from the stated axioms plus a standard seed.

full rationale

Theorem 1 is proved by an explicit regulated free-field construction: two active null pencils carrying identical non-rigid seeds, spectator pencils realizing a classical block decomposition that implements either exclusive-or (γ>0) or positive-correlation (γ<0) activation patterns, and the matched cq-conditioning rule (Eqs. 6/23–28) that converts the block weights into a log-sum-exp (or its affine limit). The resulting finite difference Δ₁₂D_γ is then shown by elementary algebra (Eqs. 38–39 or concavity of g_γ) to be strictly negative. The only external inputs are the standard axioms of the Rényi class (DPI, faithfulness, tensor additivity, matched cq conditioning) taken from the literature and the existence of a finite non-rigid one-pencil seed, which is argued from DPI under restriction plus a bounded local unitary (with a concrete chiral-current example from Longo). The authors’ own prior diagonal RQNEC result is cited only for motivation and is logically independent of the off-diagonal obstruction. Nothing is fitted, nothing is defined in terms of the target inequality, and no uniqueness or ansatz is smuggled from self-citation. The derivation is therefore self-contained against its stated assumptions.

Axiom & Free-Parameter Ledger

0 free parameters · 6 axioms · 0 invented entities

The central no-go rests on four standard properties of Rényi divergences (faithfulness, DPI, tensor additivity, matched cq conditioning) taken from the quantum-information literature, plus the existence of non-rigid free-field seeds that follows from DPI and locality. No free parameters are fitted and no new physical entities are postulated.

axioms (6)
  • domain assumption Data-processing inequality for the family D_γ
    Invoked to guarantee that one-pencil divergences are non-increasing under restriction to later cuts; standard for Petz/sandwiched/α-z Rényi divergences in the stated range.
  • domain assumption Tensor additivity D_γ(⊗ψ_i ∥ ⊗ω_i) = Σ D_γ(ψ_i ∥ ω_i)
    Used to factor the multi-pencil state; listed as Eq. (5) and required by the axioms of Müller-Lennert et al. and Tomamichel.
  • domain assumption Matched classical-quantum conditioning (log-sum-exp for γ ≠ 0, ordinary average for γ = 0)
    Eqs. (6)–(7) and End Matter; the key algebraic property that produces the classical covariance term controlling the mixed variation.
  • domain assumption Faithfulness: D_γ(ψ∥ω) = 0 iff ψ = ω
    Used to guarantee that a local unitary excitation produces a strictly positive divergence on intermediate cuts.
  • domain assumption Existence of a finite non-rigid one-pencil seed
    Assumed in Theorem 1; justified by DPI plus a bounded local unitary (or explicit chiral-current HSMI example).
  • domain assumption Null factorization of free-field Hilbert space into independent transverse pencils (up to zero modes)
    Standard null-quantization setup (Wall, Bousso et al.); allows the regulated finite-pencil counter-example to be embedded in QFT.

pith-pipeline@v1.1.0-grok45 · 14878 in / 2598 out tokens · 140282 ms · 2026-07-10T17:53:13.558245+00:00 · methodology

0 comments
read the original abstract

The quantum focusing conjecture is a mathematical expression of the idea that semiclassical gravity remains universally attractive. Its off-diagonal part is a monotonicity condition on the double null shape variation of relative entropy on distinct null generators, and has been argued to follow from strong subadditivity of entanglement entropy. Recent proof of a diagonal R\'enyi quantum null energy condition raises the question: does a full R\'enyi focusing statement also hold? We answer this question negatively for any R\'enyi-type divergence satisfying data processing, tensor additivity, and matched classical--quantum conditioning.

Figures

Figures reproduced from arXiv: 2607.07799 by Pratik Roy, Tanay Kibe.

Figure 1
Figure 1. Figure 1: FIG. 1. (a) Penrose diagram of Minkowski spacetime with a Cauchy surface [PITH_FULL_IMAGE:figures/full_fig_p003_1.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

35 extracted references · 35 canonical work pages · 27 internal anchors

  1. [1]

    A Quantum Focussing Conjecture

    R. Bousso, Z. Fisher, S. Leichenauer, and A. C. Wall, Phys. Rev. D93, 064044 (2016), arXiv:1506.02669 [hep- th]

  2. [2]

    Recovering the QNEC from the ANEC

    F. Ceyhan and T. Faulkner, Commun. Math. Phys.377, 999 (2020), arXiv:1812.04683 [hep-th]

  3. [3]

    Hollands and R

    S. Hollands and R. Longo, Commun. Math. Phys.406, 269 (2025), arXiv:2503.04651 [hep-th]

  4. [4]

    Holographic Proof of the Quantum Null Energy Condition

    J. Koeller and S. Leichenauer, Phys. Rev. D94, 024026 (2016), arXiv:1512.06109 [hep-th]

  5. [5]

    A General Proof of the Quantum Null Energy Condition

    S. Balakrishnan, T. Faulkner, Z. U. Khandker, and H. Wang, JHEP09, 020 (2019), arXiv:1706.09432 [hep- th]

  6. [6]

    H. J. Borchers, Commun. Math. Phys.143, 315 (1992)

  7. [7]

    H. W. Wiesbrock, Commun. Math. Phys.157, 83 (1993), [Erratum: Commun.Math.Phys. 184, 683–685 (1997)]

  8. [8]

    H. J. Borchers, Ann. Inst. H. Poincare Phys. Theor.63, 331 (1995)

  9. [9]

    Borchers, Commun

    H.-J. Borchers, Commun. Math. Phys.179, 703 (1996)

  10. [10]

    Extension of the structure theorem of Borchers and its application to half-sided modular inclusions

    H. Araki and L. Zsido, Rev. Math. Phys.17, 491 (2005), arXiv:math/0412061

  11. [11]

    Gravity and the Crossed Product

    E. Witten, JHEP10, 008 (2022), arXiv:2112.12828 [hep- th]

  12. [12]

    Large N algebras and generalized entropy

    V. Chandrasekaran, G. Penington, and E. Witten, JHEP 04, 009 (2023), arXiv:2209.10454 [hep-th]

  13. [13]

    Subregion algebras in classical and quantum gravity,

    V. Chandrasekaran and É. É. Flanagan, (2026), arXiv:2601.07915 [hep-th]

  14. [14]

    Constraining Quantum Fields using Modular Theory

    N. Lashkari, JHEP01, 059 (2019), arXiv:1810.09306 [hep- th]

  15. [15]

    On quantum Renyi entropies: a new generalization and some properties

    M. Müller-Lennert, F. Dupuis, O. Szehr, S. Fehr, and M. Tomamichel, J. Math. Phys.54, 122203 (2013), arXiv:1306.3142 [quant-ph]

  16. [16]

    M. M. Wilde, A. Winter, and D. Yang, Commun. Math. Phys.331, 593 (2014), arXiv:1306.1586 [quant-ph]

  17. [17]

    R\'enyi divergences as weighted non-commutative vector valued $L_p$-spaces

    M. Berta, V. B. Scholz, and M. Tomamichel, Annales Henri Poincare19, 1843 (2018), arXiv:1608.05317 [math- ph]

  18. [18]

    R\'enyi relative entropies and noncommutative $L_p$-spaces

    A. Jencova, Annales Henri Poincare19, 2513 (2018), arXiv:1609.08462 [quant-ph]

  19. [19]

    R\'enyi relative entropies and noncommutative $L_p$-spaces II

    A. Jenčová, Annales Henri Poincare22, 3235 (2021), arXiv:1707.00047 [quant-ph]

  20. [20]

    A Renyi Quantum Null Energy Condition: Proof for Free Field Theories

    M. Moosa, P. Rath, and V. P. Su, JHEP01, 064 (2021), arXiv:2007.15025 [hep-th]

  21. [21]

    Proof of Renyi QNEC for free fermions

    P. Roy, Phys. Rev. D108, 045010 (2023), arXiv:2212.02331 [hep-th]

  22. [22]

    A general proof of integer R\'enyi QNEC

    T. Kibe and P. Roy, (2026), arXiv:2605.15272 [hep-th]

  23. [23]

    K. M. R. Audenaert and N. Datta, J. Math. Phys.56, 022202 (2015), arXiv:1310.7178 [quant-ph]

  24. [24]

    A remark on non-commutative $L^p$-spaces

    S. Kato and Y. Ueda, Studia Mathematica275, 235 (2023), arXiv:2307.01790 [math.OA]

  25. [25]

    S. Kato, J. Math. Phys.65, 042202 (2024), arXiv:2311.01748 [math.OA]

  26. [26]
  27. [27]

    Quantum Information Processing with Finite Resources -- Mathematical Foundations

    M. Tomamichel,Quantum Information Processing with Finite Resources. Mathematical Foundations, Springer- Briefs in Mathematical Physics, Vol. 5 (Springer, 2016) arXiv:1504.00233 [quant-ph]

  28. [28]

    A. C. Wall, Phys. Rev. D85, 104049 (2012), [Erratum: Phys.Rev.D 87, 069904 (2013)], arXiv:1105.3445 [gr-qc]

  29. [29]

    Proof of the Quantum Null Energy Condition

    R. Bousso, Z. Fisher, J. Koeller, S. Leichenauer, and A. C. Wall, Phys. Rev. D93, 024017 (2016), arXiv:1509.02542 [hep-th]

  30. [30]

    T. A. Malik and R. Lopez-Mobilia, Phys. Rev. D101, 066028 (2020), arXiv:1910.07594 [hep-th]

  31. [31]

    Petz, Commun

    D. Petz, Commun. Math. Phys.105, 123 (1986)

  32. [32]

    Preservation of a quantum Renyi relative entropy implies existence of a recovery map

    A. Jencova, J. Phys. A50, 085303 (2017), arXiv:1604.02831 [quant-ph]

  33. [33]

    Entropy distribution of localised states

    R. Longo, Commun. Math. Phys.373, 473 (2019), arXiv:1809.03358 [hep-th]

  34. [34]

    V. P. Belavkin and P. Staszewski, Annales De L Institut Henri Poincare-physique Theorique37, 51 (1982)

  35. [35]

    The $\alpha \to 1$ Limit of the Sharp Quantum R\'enyi Divergence

    B. Bergh, R. Salzmann, and N. Datta, J. Math. Phys. 62, 092205 (2021), arXiv:2102.06576 [quant-ph]