arxiv: 2605.11024 · v1 · submitted 2026-05-10 · 🪐 quant-ph · math-ph· math.MP

Recognition: 2 theorem links

· Lean Theorem

Midpoint BKM Estimates and Boundary Coherence

Hassan Nasreddine

Authors on Pith no claims yet

Pith reviewed 2026-05-13 06:00 UTC · model grok-4.3

classification 🪐 quant-ph math-phmath.MP

keywords quantum relative entropyBKM kernelblock matricescoherence blockspectral gapHessian midpoint estimatenoncommutative boundsFrobenius norm

0 comments

The pith

A midpoint BKM Hessian estimate yields a noncommutative lower bound on quantum relative entropy to the block-diagonal part that simplifies to an explicit logarithmic form proportional to the squared Frobenius norm of the off-diagonal block.

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

The paper establishes lower bounds on the quantum relative entropy between a density matrix and its block-diagonal projection using the Bogoliubov-Kubo-Mori kernel. The bound is derived from a midpoint estimate of the BKM Hessian along the straight-line interpolation from the full matrix to its block-diagonal version. This approach works in the noncommutative setting and preserves details of the joint spectrum of the blocks and the coherence term. Under an additional spectral gap between the diagonal blocks, the bound becomes an explicit logarithm of one plus a multiple of the squared Frobenius norm of the off-diagonal block. Such estimates matter for quantifying how much quantum coherence or information is lost when discarding off-diagonal terms without assuming the blocks commute.

Core claim

For a block matrix with positive diagonal blocks A and C and off-diagonal coherence block B, a lower bound on the quantum relative entropy to the block-diagonal part is proved using the associated BKM kernel. The proof relies on a midpoint estimate for the BKM Hessian along the affine interpolation path. Under a spectral gap condition on A relative to C, this yields an explicit logarithmic lower bound proportional to the squared Frobenius norm of B. The BKM metric appears naturally because it coincides with the Hessian of quantum relative entropy.

What carries the argument

The midpoint estimate for the BKM Hessian along the affine interpolation between the density matrix and its block-diagonal projection.

If this is right

The lower bound is genuinely noncommutative and retains information about the joint spectral structure of the diagonal blocks and the coherence term.
Under a spectral gap condition on A relative to C, the bound simplifies to an explicit logarithmic form proportional to the squared Frobenius norm of B.
The BKM metric enters naturally because it is the Hessian of the quantum relative entropy functional.
The estimate applies directly to density matrices without requiring the diagonal blocks to commute.

Where Pith is reading between the lines

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

The bound offers a concrete way to quantify residual coherence after projection onto block-diagonal states in systems with natural spectral separation between blocks.
Similar midpoint Hessian techniques might extend to other quantum divergence functionals or to multipartite block structures beyond two blocks.
Numerical checks on low-dimensional qubit or qutrit examples with tunable gaps could directly verify the scaling with the Frobenius norm of B.

Load-bearing premise

The spectral gap condition on A relative to C together with the assumption that the BKM Hessian midpoint estimate directly controls the relative entropy difference without additional commutativity or positivity violations.

What would settle it

A concrete block matrix example where the quantum relative entropy to the block-diagonal part falls below the BKM-kernel lower bound, or where the explicit logarithmic form fails to hold despite the spectral gap condition on A relative to C being satisfied.

read the original abstract

We study lower bounds for the quantum relative entropy between a density matrix and its block-diagonal part. For a block matrix with diagonal blocks A,C>0 and off-diagonal coherence block B, we prove a lower bound expressed through the associated Bogoliubov--Kubo--Mori (BKM) kernel. The proof uses a midpoint estimate for the BKM Hessian along the affine interpolation between the matrix and its block-diagonal projection. The resulting estimate is genuinely noncommutative and retains information about the joint spectral structure of the diagonal blocks and the coherence term. As a consequence, under a spectral gap condition on A relative to C, we obtain an explicit logarithmic lower bound proportional to the squared Frobenius norm of the coherence block. The appearance of the BKM metric is natural in this setting because it coincides with the Hessian of quantum relative entropy.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

Midpoint BKM Hessian estimate for relative entropy to block-diagonal projection is the new piece, but the step that turns it into a lower bound on the integral needs a monotonicity argument that the abstract does not supply.

read the letter

The paper's main contribution is a midpoint estimate on the BKM Hessian along the affine path from a block matrix to its diagonal projection. This produces a lower bound on the relative entropy that keeps the off-diagonal coherence block and the joint spectrum of the diagonal blocks in view, rather than discarding them. Under a spectral gap between the diagonal blocks, the bound simplifies to something proportional to the squared Frobenius norm of the coherence term times a log factor. That explicit form is the part that could actually get used in resource-theory calculations or mixing-time estimates for block-structured systems.

Referee Report

2 major / 2 minor

Summary. The manuscript claims a lower bound on the quantum relative entropy S(ρ || D) for a block matrix ρ with positive diagonal blocks A, C and off-diagonal coherence block B, where D is the block-diagonal projection of ρ. The bound is obtained via a midpoint estimate on the BKM Hessian along the affine path ρ_t = (1-t)D + tρ, yielding an expression in terms of the associated BKM kernel; under a spectral gap condition on A relative to C, this specializes to an explicit logarithmic lower bound proportional to the squared Frobenius norm of B. The proof strategy rests on the fact that the BKM metric is the Hessian of the relative entropy.

Significance. If the midpoint estimate is rigorously justified, the result supplies a genuinely noncommutative lower bound on relative entropy that retains joint spectral information between the diagonal blocks and the coherence term. This is a natural application of the Hessian property and could be useful for quantitative coherence measures or boundary estimates in quantum information. The derivation is parameter-free in the sense described and directly exploits the integral representation of relative entropy.

major comments (2)

[proof of the midpoint BKM estimate (likely §3)] The central lower-bound claim (abstract and the proof of the main estimate) replaces the integrand in the integral representation ∫_0^1 (1-t) ⟨ρ-D, Hess_{ρ_t}(ρ-D)⟩ dt by its value at the midpoint t=1/2. No explicit argument is supplied showing that the quadratic form t ↦ ⟨X, Hess_{ρ_t} X⟩ is minimized (or bounded below) at t=1/2 for non-commuting A and C; without monotonicity or convexity of this map along the path, the midpoint value can overestimate rather than underestimate the integral, invalidating the claimed lower bound. This is load-bearing for the entire result.
[statement and proof of the explicit logarithmic bound] The spectral-gap condition on A relative to C is invoked only after the midpoint estimate to extract the logarithmic factor. If the midpoint step already fails to produce a valid lower bound, the subsequent specialization cannot be salvaged without first repairing the integral comparison.

minor comments (2)

[abstract and §2] Clarify the precise definition of the BKM kernel used in the final bound and its relation to the standard BKM metric; the notation shifts between 'BKM kernel' and 'BKM Hessian' without an explicit cross-reference.
[preliminaries] The affine interpolation ρ_t is defined with D the block-diagonal projection, but the positivity of ρ_t for t ∈ [0,1] is not verified when A and C have different spectra; add a short remark on the domain of the Hessian.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading and for identifying the need for an explicit justification of the midpoint estimate. We address the major comments point by point below.

read point-by-point responses

Referee: [proof of the midpoint BKM estimate (likely §3)] The central lower-bound claim (abstract and the proof of the main estimate) replaces the integrand in the integral representation ∫_0^1 (1-t) ⟨ρ-D, Hess_{ρ_t}(ρ-D)⟩ dt by its value at the midpoint t=1/2. No explicit argument is supplied showing that the quadratic form t ↦ ⟨X, Hess_{ρ_t} X⟩ is minimized (or bounded below) at t=1/2 for non-commuting A and C; without monotonicity or convexity of this map along the path, the midpoint value can overestimate rather than underestimate the integral, invalidating the claimed lower bound. This is load-bearing for the entire result.

Authors: We agree that the current manuscript does not supply an explicit argument establishing that the quadratic form t ↦ ⟨X, Hess_{ρ_t} X⟩ attains its minimum at t=1/2. The derivation implicitly relies on properties of the BKM kernel, but this step requires a dedicated justification. In the revised version we will add a lemma proving the requisite lower bound on the integrand, using the integral representation of the BKM inner product and the operator convexity of the relevant map along the affine path. This will rigorously support the midpoint replacement and the claimed lower bound on the relative entropy. revision: yes
Referee: [statement and proof of the explicit logarithmic bound] The spectral-gap condition on A relative to C is invoked only after the midpoint estimate to extract the logarithmic factor. If the midpoint step already fails to produce a valid lower bound, the subsequent specialization cannot be salvaged without first repairing the integral comparison.

Authors: We concur that the explicit logarithmic bound is conditional on the validity of the midpoint estimate. Once the supporting lemma for the midpoint step is included, the specialization under the spectral-gap assumption on the diagonal blocks follows directly from the properties of the BKM kernel and will be presented with a clear logical sequence in the revised manuscript. revision: yes

Circularity Check

0 steps flagged

No circularity: derivation uses standard Hessian property and midpoint estimate as independent steps

full rationale

The paper states that the BKM metric coincides with the Hessian of quantum relative entropy (a known external fact) and then applies a midpoint estimate along the affine interpolation ρ_t = (1-t)D + tρ to bound the relative entropy integral. No equation reduces the final lower bound to an input quantity by construction, no parameters are fitted and then relabeled as predictions, and no self-citation chain or ansatz is invoked to justify the central steps. The spectral-gap assumption appears only for the explicit logarithmic form and does not close a loop. The argument is therefore self-contained against external mathematical benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The central claim rests on standard properties of the quantum relative entropy and the BKM metric (its Hessian character) plus domain assumptions on positivity and the spectral gap; no free parameters or new entities are introduced.

axioms (2)

domain assumption A and C are positive definite (A,C>0)
Required for the density matrix to be valid and for the relative entropy and BKM kernel to be well-defined.
domain assumption Spectral gap condition on A relative to C
Invoked to convert the general BKM bound into the explicit logarithmic form proportional to ||B||_F^2.

pith-pipeline@v0.9.0 · 5436 in / 1452 out tokens · 84746 ms · 2026-05-13T06:00:58.280697+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

D(ρ∥Πρ)≥Tr[B*Ω^{-1}_{A,C}(B)] where Ω^{-1}_{A,C}(B)=∫_0^∞(A+rI)^{-1}B(C+rI)^{-1}dr; midpoint estimate HM+tY(Y,Y)≥HM(Y,Y) from block involution UY U^*=−Y
IndisputableMonolith/Foundation/AlphaCoordinateFixation.lean J_uniquely_calibrated_via_higher_derivative unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

logarithmic boundary law D(ρ∥Πρ)≥∥B∥_F² log(λ_min(A)/Tr C) under spectral-gap condition

What do these tags mean?

matches: The paper's claim is directly supported by a theorem in the formal canon.
supports: The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends: The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses: The paper appears to rely on the theorem as machinery.
contradicts: The paper's claim conflicts with a theorem or certificate in the canon.
unclear: Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

20 extracted references · 20 canonical work pages · 1 internal anchor

[1]

Baumgratz, M

T. Baumgratz, M. Cramer, and M. B. Plenio, Quantifying coherence,Phys. Rev. Lett. 113(2014), 140401

work page 2014
[2]

N. N. Bogoliubov, On a variational principle in the many-body problem,Sov. Phys. Dokl.3(1958), 292–294

work page 1958
[3]

E. A. Carlen and J. Maas, Gradient flow and entropy inequalities for quantum Markov semigroups with detailed balance,J. Funct. Anal.273(2017), 1810–1869

work page 2017
[4]

Hiai and D

F. Hiai and D. Petz, The proper formula for relative entropy and its asymptotics in quantum probability,Commun. Math. Phys.143(1991), 99–114

work page 1991
[5]

M. J. Kastoryano and K. Temme, Quantum logarithmic Sobolev inequalities and rapid mixing,J. Math. Phys.54(2013), 052202

work page 2013
[6]

Kubo, Statistical-mechanical theory of irreversible processes

R. Kubo, Statistical-mechanical theory of irreversible processes. I,J. Phys. Soc. Japan 12(1957), 570–586

work page 1957
[7]

Lesniewski and M

A. Lesniewski and M. B. Ruskai, Monotone Riemannian metrics and relative operator entropy,J. Math. Phys.40(1999), 5702–5724

work page 1999
[8]

Mori, Transport, collective motion, and Brownian motion,Prog

H. Mori, Transport, collective motion, and Brownian motion,Prog. Theor. Phys.33 (1965), 423–455

work page 1965
[9]

M¨ uller-Hermes and D

A. M¨ uller-Hermes and D. S. Fran¸ ca, Sandwiched R´ enyi convergence for quantum evo- lutions,Quantum2(2018), 55

work page 2018
[10]

Ohya and D

M. Ohya and D. Petz,Quantum Entropy and Its Use, Springer, Berlin, 1993

work page 1993
[11]

Entropy Moduli and Support-Sensitive BKM Coercivity for Rank-Deficient Non-Commutative Markov Semigroups

H. Nasreddine,Entropy Moduli and Support-Sensitive BKM Coercivity for Rank- Deficient Non-Commutative Markov Semigroups, arXiv:2604.16616v2 (2026)

work page internal anchor Pith review Pith/arXiv arXiv 2026
[12]

Accardi and C

L. Accardi and C. Cecchini,Conditional expectations in von Neumann algebras and a theorem of Takesaki, J. Funct. Anal.45(1982), 245–273

work page 1982
[13]

Kubo and T

F. Kubo and T. Ando,Means of positive linear operators, Math. Ann.246(1980), 205–224

work page 1980
[14]

Petz, Quasi-entropies for finite quantum systems,Rep

D. Petz, Quasi-entropies for finite quantum systems,Rep. Math. Phys.23(1986), 57–65

work page 1986
[15]

Petz and C

D. Petz and C. Sudar, Geometries of quantum states,J. Math. Phys.37(1996), 2662– 2673

work page 1996
[16]

Streltsov, G

A. Streltsov, G. Adesso, and M. B. Plenio, Colloquium: Quantum coherence as a re- source,Rev. Mod. Phys.89(2017), 041003

work page 2017
[17]

Vedral, The role of relative entropy in quantum information theory,Rev

V. Vedral, The role of relative entropy in quantum information theory,Rev. Mod. Phys. 74(2002), 197–234. 23

work page 2002
[18]

Winter and D

A. Winter and D. Yang, Operational resource theory of coherence,Phys. Rev. Lett.116 (2016), 120404

work page 2016
[19]

Wirth and H

M. Wirth and H. Zhang, Complete gradient estimates of quantum Markov semigroups, Commun. Math. Phys.387(2021), 343–401

work page 2021
[20]

M. A. Nielsen and I. L. Chuang,Quantum Computation and Quantum Information, Cambridge University Press, 2000. 24

work page 2000