Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:2605.11024

Sharp Entropy Cost of Coherent Leakage Across a Block Decomposition

Hassan Nasreddine

quant-phmath-phSubmitted May 10, 2026• Announced May 12, 2026• 0 citations• Revised Aug 9, 2026 (v2)
v1 · May 10, 2026v2 · Aug 9, 2026
arXivHTMLPDF

Abstract

Let $\mathcal{H}=P\mathcal{H}\oplus Q\mathcal{H}$ and let $Πρ=PρP\oplus QρQ$ be the two-block pinching of a density matrix. Write $A=PρP$, $B=PρQ$, and $C=QρQ$. We study the block-coherence entropy $D(ρ\|Πρ)$, the relative-entropy cost of removing cross-block coherence. First, we prove a midpoint Bogoliubov--Kubo--Mori coercivity estimate, giving the operator-kernel lower bound $D(ρ\|Πρ)\ge \operatorname{Tr}[B^*Ω^{-1}_{A,C}(B)]$ for strictly positive diagonal blocks. The proof uses the block-sign involution $U=P-Q$ to obtain a sign-symmetric midpoint lower bound for the BKM Hessian. Second, we determine the exact minimum in the associated constrained entropy-minimization problem: at fixed protected-sector floor, leakage mass, and quadratic coherence, the minimum can be achieved by a single active two-level block, independently of the ambient dimension. Third, under a uniform floor $A\ge a_0P$ and $0<\varepsilon_Q\le a_0/2$, we obtain the logarithmic boundary law $D(ρ\|Πρ)\ge \|B\|_F^2\log(a_0/\varepsilon_Q)$, showing that the lower-bound coefficient multiplying the quadratic coherence diverges logarithmically as leakage mass tends to zero. As a finite-dimensional application, the estimates give an entropy certificate for coherent leakage from a protected subspace: $PρQ$ measures protected--leakage coherence, and the constrained entropy minimum is attained by a state supported on a single effective two-level transition. The BKM estimate also yields a dissipation lower bound along an idealized uniform block-dephasing orbit. Finally, the midpoint mechanism extends to finite-dimensional trace-preserving $\mathbb{Z}_2$-graded automorphisms and bounded finite-trace semifinite corners.

Comments
26 pages, no figures. Revised and expanded version; title changed to reflect the broader scope. Clarified several proofs and boundary formulations, added references and provenance, and expanded the variational, application, graded, and semifinite discussions. The principal finite-dimensional results are unchanged

Related papers

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

Hassan Nasreddine

We study entropy--coherence relations near rank-deficient support boundaries in finite-dimensional quantum systems. For block-diagonal reference states, we establish support-sensitive coercivity estimates showing that the entropy cost of cross-boundary coherence acquires a logarithmic enhancement as the population scale approaches the support boundary. Combined with finite-time entropy bounds, these estimates yield conditional entropy--activation bounds with a logarithmic correction factor of order \(e^{-αt}(1+αt)^{-1/2}\) in coherence-dominant regimes. The analysis proceeds through pinching reductions and effective \(2\times2\) Bogoliubov--Kubo--Mori (BKM) estimates adapted to the coherence--population structure. We further apply the framework to Davies semigroups under additional secular decoupling and population-rate assumptions. The resulting statements provide conditional certification bounds near rank-deficient stationary states, rather than general mixing-time or convergence-rate estimates.

quant-pharXiv:2604.16616•Apr 21, 2026•1 citation

Sharp dimension-independent Schatten bounds for PPT-protected coherence

Zixuan He

How much coherence can remain when its two protecting diagonal blocks have positive partial transpose? This article gives a sharp bound without fixing the local Hilbert-space dimensions. Let H be a positive two-by-two operator block matrix with diagonal blocks A and B and off-diagonal block X. Each diagonal block is bipartite and has positive partial transpose internally. The coherence X may be rectangular, complex and non-Hermitian. Define two crossed arrangements by \(\mathcal R(X)_{ik,j\ell}=X_{ij,k\ell}\) and \(\mathcal G(X)_{i\ell,kj}=X_{ij,k\ell}\). Then \[\max\{\|\mathcal R(X)\|_{4/3},\|\mathcal G(X)\|_{4/3}\}\le\sqrt{\operatorname{Tr}A\,\operatorname{Tr}B}.\] The traces and Schatten norms are unnormalized. The constant one is optimal, and every smaller Schatten exponent fails uniformly in dimension. Internal PPT protection of the diagonal blocks is distinct from requiring the entire flagged block matrix to be PPT. Why the endpoint is sharp. Marginal spectral truncation gives a dimension-independent singular-value tail of order \(n^{-3/4}\). Closure under tensor powers removes the prefactor and returns the strong \(4/3\) endpoint. A Fourier shield family already present in the private-state literature supplies matching equality and the obstruction below the endpoint; the new contribution is the universal upper bound and its endpoint identification. Infinite-dimensional use. The paper extends the result to trace-class guards on separable Hilbert spaces, specifies the partial-transpose and local-filter domains, and states the joint weighted-trace conditions for unbounded filters. Trace-norm conversion and partial trace incur additional explicit costs. The integrated 10-page article includes two figure groups, complete proofs and references.

•Nov 8, 2026•0 citations

Entropy Stability under Block-Separable Constraints: Fixed-Weight Rank-Lattice Envelopes and Convexification

Hassan Nasreddine

For block-diagonal quantum states with prescribed block weights and dimensions, we determine the minimum von Neumann entropy excess at each trace-norm distance from the blockwise-pure states. The one-block extremal is the classical reverse-Fano profile; the new problem is its constrained assembly when the block weights and dimensions cannot be redistributed. The resulting value function $Φ_{q,n}(D)$ is attained on the full admissible interval and is the lower envelope of finitely many rank-lattice edges. Every minimizing distance allocation has at most one block between consecutive integer ranks, and the active edge can switch nonmonotonically with $D$. We also determine the lower convex envelope of $Φ_{q,n}$ explicitly: it is the polygonal curve through synchronized rank vertices and admits a one-parameter dual representation. The exact nonconvex envelope differs from this convex surrogate by at most $q_\diamond\log(5/4)$, with a sharper stage-dependent bound, and the constant is optimal. For binomial charge-sector weights the uniform gap is $O(L^{-1/2})$. The framework yields a symmetry-sector purity certificate; in a three-qubit phase-damping example the fixed-weight certificate remains nontrivial where the universal classical entropy-error envelope is already vacuous.

quant-pharXiv:2512.16192•Dec 18, 2025•1 citation

Sharp entropy thresholds and spectral dispersion in coherent-state approximation

Zixuan He

When can a finite phase-space entropy cease to constrain an optical state's distance from every coherent state? This article gives the earliest possible boundary over all normal optical sources and explains how the distribution of their eigenvalues determines that boundary. It is intended for researchers using entropy to justify coherent-state approximation when the number of modes can grow. For an \(m\)-mode state, let \(Q_\rho(z)=\langle z|\rho|z\rangle\), \(d\mu_m(z)=\pi^{-m}d^{2m}z\), and define \[M_p(\rho)=p^m\int Q_\rho(z)^p\,d\mu_m(z),\qquad \delta_p(\rho)=-\frac{\log M_p(\rho)}{p-1}.\] The approximation error is the full trace-norm distance \(D(\rho)=\inf_\alpha\|\rho-|\alpha\rangle\langle\alpha|\|_1\), with maximum two. Write \(F_p(s)\) for the supremum of \(D^2\) at exact excess \(s\), over all finite mode numbers, and \(F_p^{\mathrm{pure}}(s)\) for the pure-state version. Exact all-source thresholds. For each fixed real order \(p\ge2\), \[F_p(s)=4\iff s\ge\frac{p\log p}{p-1},\qquad F_p^{\mathrm{pure}}(s)=4\iff s\ge\frac{2p\log p-\log\Gamma(p+1)}{p-1}.\] At second order the thresholds are \(\log4\) and \(\log8\). Below the respective threshold there is a positive distance gap from two, uniform in mode number. Above it, physical sequences attain the supremum at exactly fixed entropy. Each individual source has finitely many modes; arbitrary complex coherence and infinite photon tails are allowed, without an input-energy bound. The maximum is a limiting supremum, not the distance of an individual normal state. Spectral dispersion explains the pure/mixed difference. Along a sequence becoming maximally distant, let the decreasing eigenvalues converge to \(\ell_j\), and set \(\theta=1-\sum_j\ell_j\). This is spectral mass spread among vanishing eigenvalues, not probability discarded from the source. With independent mean-one exponential variables \(E_j\), the sharp limiting moment capacity is \[\mathcal A_p(\ell)=p^{-2p}\mathbb E\!\left(\sum_j\ell_jE_j+p\theta\right)^p.\] Every limiting spectrum has a physical attaining sequence. At second order, \(\mathcal A_2=[(1+\theta)^2+\sum_j\ell_j^2]/16\), which gives the exact minimum dispersed mass \[\theta_{\min}(s)=\sqrt{8e^{-s}-1},\qquad \log4\le s\le\log8.\] Quantitative refinements. The article retains explicit bounds and one-photon probability rigidity for the stated number-diagonal and coherent-label source classes, minimum limiting photon costs of one for mixed sources and two for pure sources, and the sharp number-diagonal distance slope \(4/\sqrt e\). These finer source-class conclusions are kept distinct from the universal plateau theorem. At second order, the moment is also half-loss output purity and a two-copy interference observable. Version 2.0 extends the earlier source-class analysis to the full-source earliest platforms and limiting-spectrum capacity. The integrated 28-page PDF includes complete proofs and three figure groups. It determines the plateau onsets and spectral capacities.

•Oct 31, 2026•0 citations

Exact Reference-Assisted Entropy Contraction from the Choi Matrix

Harjas Sandhu

How much distinguishability can a noisy quantum channel preserve when its input is correlated with an untouched quantum memory? This paper gives an exact finite semidefinite program for that question for every finite-dimensional completely positive, trace-preserving channel. The complete coefficient compares joint states with the same memory marginal. The conditional coefficient instead subtracts the distinguishability already present in the memory. Both coefficients, and conditional mutual-information contraction with a separate conditioning system, equal one minus the reciprocal of an optimization in the channel’s unnormalized Choi matrix. The equivalent inverse formula retains an essential range condition, including for singular channels. The proof solves a centered matrix-Gram problem, shows that its optimization center can be positive, and constructs a signed observable lift through an erasure channel. Two weighted quadratic inequalities give the universal entropy bound. A rare coherent memory sector proves sharpness while preserving both local marginals exactly. A memory of bounded dimension suffices, including with prescribed faithful input marginals. The same program gives exact coefficients for regular modular Petz divergences whose generators have finite endpoint value and recession slope, every fixed Petz Rényi order strictly between zero and one, and quantum Chernoff divergence. Further results include a harmonic dual, an exact tensor-product law, a strict-contraction criterion, a sharp comparison with positive quantum Doeblin minorization, and explicit classical, depolarizing and orthogonal-unitary-noise formulas. Ordinary quantum mutual-information contraction is a different optimization. For qubit depolarization the paper determines its exact value on the full completely positive parameter interval. At shrinkage one half, ordinary relative-entropy, ordinary mutual-information and complete entropy contraction are respectively 1/4, 1/3 and 3/8. In higher dimension, bounds and a strict separation are proved without asserting an exact ordinary-MI formula. A discontinuity between nearly complete amplitude damping and pure reset rules out uniformly bounded-use estimation of the exact coefficient from an unknown channel, even with adaptive experiments and quantum ancillas. Thus evaluation from exact supplied Choi data does not imply stable experimental acquisition. Rare sharpness witnesses can carry vanishing absolute information; the results do not assert a communication rate, physical simulator, general recovery map or polynomial cost in physical qubit count.

•Nov 14, 2026•0 citations

Hidden quantum coherence at the Wigner entropy endpoint

Zixuan He

How much high-photon quantum coherence is controlled by a nonnegative phase-space distribution? This article gives quantitative guarantees for approximating an optical state by a finite photon-number block, and constructs physical states showing why the guarantee cannot be inferred from mean energy or Wigner positivity alone at the strength available for a Gaussian-mixture source. Source classes and scope. The comparison concerns one optical mode with \([Q,P]=i\), mean photon number at most \(E>0\), and a known parity centre. The full class \(\mathcal W_E\) consists of normal Wigner-positive states at the exact vacuum min-entropy endpoint, equivalently states supported on the even photon sector in that frame. The subclass \(\mathcal G_E\) consists of actual probability mixtures of centred pure squeezed vacua. Mixture labels are unavailable, component energies may be arbitrarily large, and their true average energy is charged. Infinite photon tails are retained. A hierarchy of coherent-tail guarantees. Let \(P_L\) project onto photon numbers at most \(L\), and define the cutoff error by \[e_L(\rho)=\tfrac12\|\rho-P_L\rho P_L\|_1.\] This quantity includes both discarded photon population and coherence across the cutoff. A mean-energy constraint alone gives the standard \(O_E(L^{-1/2})\) guarantee. At the exact Wigner endpoint, the article proves \[\sup_{\rho\in\mathcal W_E}e_L(\rho)\le C_E(L+1)^{-3/4},\qquad \sup_{\rho\in\mathcal G_E}e_L(\rho)\le \frac{4E}{L+1},\qquad L\ge2.\] The stronger endpoint bound combines operator positivity and Wigner positivity on the same density kernel, followed by a controlled conversion to trace norm. It requires only mean energy, without a uniform second photon moment. Physical coherence beyond the Gaussian-mixture bound. A family \(\tau_d\in\mathcal W_E\) has exact energy \(E\), covariance \((E+1/2)I\), strictly positive Wigner functions and nonzero population at every even photon number. For cutoffs \(L_d=\Theta(d^3)\), \[e_{L_d}(\tau_d)\ge c_Ed^{-5/2},\qquad (L_d+1)e_{L_d}(\tau_d)\longrightarrow\infty.\] Thus the Gaussian-mixture \(L^{-1}\) law fails uniformly on the full endpoint class. The construction preserves coherence through an actual optical partial trace and pays its physical energy cost. Positive phase mixing further yields distinct states with identical complete photon statistics and covariance but nearly indistinguishable bounded optical responses. The rare bright branches obey the prescribed mean-energy budget; no efficient experimental preparation protocol or uniform branch-energy bound is assumed. Consequences for a specified measurement system. Each independent copy is measured by bounded centred squeezing followed by a vacuum/nonvacuum binary test. At fixed energy, squeezing range and confidence, let \(n_X(\delta)\) be the minimum worst-case number of copies for trace-distance accuracy \(\delta\) over source class \(X\). As \(\delta\downarrow0\), the bounds are \[\log n_{\mathcal G_E}(\delta)=\Theta(\delta^{-1}),\qquad c\delta^{-6/5}\le\log n_{\mathcal W_E}(\delta)\le C\delta^{-4/3}.\] These are measurement-restricted reconstruction costs. The lower bounds allow adaptive randomized settings. The upper bounds use finite designs and estimates in the original source class, with the energy budget respected for every record; they establish statistical existence rather than an efficient catalogue-generation algorithm. The full-class exponents remain distinct lower and upper bounds. The article also characterizes infinite-order Wigner entropy equality by displaced even-parity support and corrects the pure-Gaussian-only equality clause at infinite order in Dias and Prata (2023). The 27-page integrated article includes the complete proofs, figures and references.

•Oct 26, 2026•0 citations
Let H=PH⊕QH\mathcal{H}=P\mathcal{H}\oplus Q\mathcal{H}H=PH⊕QH and let Πρ=PρP⊕QρQΠρ=PρP\oplus QρQΠρ=PρP⊕QρQ be the two-block pinching of a density matrix. Write A=PρPA=PρPA=PρP, B=PρQB=PρQB=PρQ, and C=QρQC=QρQC=QρQ. We study the block-coherence entropy D(ρ∥Πρ)D(ρ\|Πρ)D(ρ∥Πρ), the relative-entropy cost of removing cross-block coherence. First, we prove a midpoint Bogoliubov--Kubo--Mori coercivity estimate, giving the operator-kernel lower bound D(ρ∥Πρ)≥Tr⁡[B∗ΩA,C−1(B)]D(ρ\|Πρ)\ge \operatorname{Tr}[B^*Ω^{-1}_{A,C}(B)]D(ρ∥Πρ)≥Tr[B∗ΩA,C−1​(B)] for strictly positive diagonal blocks. The proof uses the block-sign involution U=P−QU=P-QU=P−Q to obtain a sign-symmetric midpoint lower bound for the BKM Hessian. Second, we determine the exact minimum in the associated constrained entropy-minimization problem: at fixed protected-sector floor, leakage mass, and quadratic coherence, the minimum can be achieved by a single active two-level block, independently of the ambient dimension. Third, under a uniform floor A≥a0PA\ge a_0PA≥a0​P and 0<εQ≤a0/20<\varepsilon_Q\le a_0/20<εQ​≤a0​/2, we obtain the logarithmic boundary law D(ρ∥Πρ)≥∥B∥F2log⁡(a0/εQ)D(ρ\|Πρ)\ge \|B\|_F^2\log(a_0/\varepsilon_Q)D(ρ∥Πρ)≥∥B∥F2​log(a0​/εQ​), showing that the lower-bound coefficient multiplying the quadratic coherence diverges logarithmically as leakage mass tends to zero. As a finite-dimensional application, the estimates give an entropy certificate for coherent leakage from a protected subspace: PρQPρQPρQ measures protected--leakage coherence, and the constrained entropy minimum is attained by a state supported on a single effective two-level transition. The BKM estimate also yields a dissipation lower bound along an idealized uniform block-dephasing orbit. Finally, the midpoint mechanism extends to finite-dimensional trace-preserving Z2\mathbb{Z}_2Z2​-graded automorphisms and bounded finite-trace semifinite corners.
We study entropy--coherence relations near rank-deficient support boundaries in finite-dimensional quantum systems. For block-diagonal reference states, we establish support-sensitive coercivity estimates showing that the entropy cost of cross-boundary coherence acquires a logarithmic enhancement as the population scale approaches the support boundary. Combined with finite-time entropy bounds, these estimates yield conditional entropy--activation bounds with a logarithmic correction factor of order e−αt(1+αt)−1/2e^{-αt}(1+αt)^{-1/2}e−αt(1+αt)−1/2 in coherence-dominant regimes. The analysis proceeds through pinching reductions and effective 2×22\times22×2 Bogoliubov--Kubo--Mori (BKM) estimates adapted to the coherence--population structure. We further apply the framework to Davies semigroups under additional secular decoupling and population-rate assumptions. The resulting statements provide conditional certification bounds near rank-deficient stationary states, rather than general mixing-time or convergence-rate estimates.
How much coherence can remain when its two protecting diagonal blocks have positive partial transpose? This article gives a sharp bound without fixing the local Hilbert-space dimensions. Let H be a positive two-by-two operator block matrix with diagonal blocks A and B and off-diagonal block X. Each diagonal block is bipartite and has positive partial transpose internally. The coherence X may be rectangular, complex and non-Hermitian. Define two crossed arrangements by R(X)ik,jℓ=Xij,kℓ\mathcal R(X)_{ik,j\ell}=X_{ij,k\ell}R(X)ik,jℓ​=Xij,kℓ​ and G(X)iℓ,kj=Xij,kℓ\mathcal G(X)_{i\ell,kj}=X_{ij,k\ell}G(X)iℓ,kj​=Xij,kℓ​. Then max⁡{∥R(X)∥4/3,∥G(X)∥4/3}≤Tr⁡A Tr⁡B.\max\{\|\mathcal R(X)\|_{4/3},\|\mathcal G(X)\|_{4/3}\}\le\sqrt{\operatorname{Tr}A\,\operatorname{Tr}B}.max{∥R(X)∥4/3​,∥G(X)∥4/3​}≤TrATrB​. The traces and Schatten norms are unnormalized. The constant one is optimal, and every smaller Schatten exponent fails uniformly in dimension. Internal PPT protection of the diagonal blocks is distinct from requiring the entire flagged block matrix to be PPT. Why the endpoint is sharp. Marginal spectral truncation gives a dimension-independent singular-value tail of order n−3/4n^{-3/4}n−3/4. Closure under tensor powers removes the prefactor and returns the strong 4/34/34/3 endpoint. A Fourier shield family already present in the private-state literature supplies matching equality and the obstruction below the endpoint; the new contribution is the universal upper bound and its endpoint identification. Infinite-dimensional use. The paper extends the result to trace-class guards on separable Hilbert spaces, specifies the partial-transpose and local-filter domains, and states the joint weighted-trace conditions for unbounded filters. Trace-norm conversion and partial trace incur additional explicit costs. The integrated 10-page article includes two figure groups, complete proofs and references.
For block-diagonal quantum states with prescribed block weights and dimensions, we determine the minimum von Neumann entropy excess at each trace-norm distance from the blockwise-pure states. The one-block extremal is the classical reverse-Fano profile; the new problem is its constrained assembly when the block weights and dimensions cannot be redistributed. The resulting value function Φq,n(D)Φ_{q,n}(D)Φq,n​(D) is attained on the full admissible interval and is the lower envelope of finitely many rank-lattice edges. Every minimizing distance allocation has at most one block between consecutive integer ranks, and the active edge can switch nonmonotonically with DDD. We also determine the lower convex envelope of Φq,nΦ_{q,n}Φq,n​ explicitly: it is the polygonal curve through synchronized rank vertices and admits a one-parameter dual representation. The exact nonconvex envelope differs from this convex surrogate by at most q⋄log⁡(5/4)q_\diamond\log(5/4)q⋄​log(5/4), with a sharper stage-dependent bound, and the constant is optimal. For binomial charge-sector weights the uniform gap is O(L−1/2)O(L^{-1/2})O(L−1/2). The framework yields a symmetry-sector purity certificate; in a three-qubit phase-damping example the fixed-weight certificate remains nontrivial where the universal classical entropy-error envelope is already vacuous.
When can a finite phase-space entropy cease to constrain an optical state's distance from every coherent state? This article gives the earliest possible boundary over all normal optical sources and explains how the distribution of their eigenvalues determines that boundary. It is intended for researchers using entropy to justify coherent-state approximation when the number of modes can grow. For an mmm-mode state, let Qρ(z)=⟨z∣ρ∣z⟩Q_\rho(z)=\langle z|\rho|z\rangleQρ​(z)=⟨z∣ρ∣z⟩, dμm(z)=π−md2mzd\mu_m(z)=\pi^{-m}d^{2m}zdμm​(z)=π−md2mz, and define Mp(ρ)=pm∫Qρ(z)p dμm(z),δp(ρ)=−log⁡Mp(ρ)p−1.M_p(\rho)=p^m\int Q_\rho(z)^p\,d\mu_m(z),\qquad \delta_p(\rho)=-\frac{\log M_p(\rho)}{p-1}.Mp​(ρ)=pm∫Qρ​(z)pdμm​(z),δp​(ρ)=−p−1logMp​(ρ)​. The approximation error is the full trace-norm distance D(ρ)=inf⁡α∥ρ−∣α⟩⟨α∣∥1D(\rho)=\inf_\alpha\|\rho-|\alpha\rangle\langle\alpha|\|_1D(ρ)=infα​∥ρ−∣α⟩⟨α∣∥1​, with maximum two. Write Fp(s)F_p(s)Fp​(s) for the supremum of D2D^2D2 at exact excess sss, over all finite mode numbers, and Fppure(s)F_p^{\mathrm{pure}}(s)Fppure​(s) for the pure-state version. Exact all-source thresholds. For each fixed real order p≥2p\ge2p≥2, Fp(s)=4  ⟺  s≥plog⁡pp−1,Fppure(s)=4  ⟺  s≥2plog⁡p−log⁡Γ(p+1)p−1.F_p(s)=4\iff s\ge\frac{p\log p}{p-1},\qquad F_p^{\mathrm{pure}}(s)=4\iff s\ge\frac{2p\log p-\log\Gamma(p+1)}{p-1}.Fp​(s)=4⟺s≥p−1plogp​,Fppure​(s)=4⟺s≥p−12plogp−logΓ(p+1)​. At second order the thresholds are log⁡4\log4log4 and log⁡8\log8log8. Below the respective threshold there is a positive distance gap from two, uniform in mode number. Above it, physical sequences attain the supremum at exactly fixed entropy. Each individual source has finitely many modes; arbitrary complex coherence and infinite photon tails are allowed, without an input-energy bound. The maximum is a limiting supremum, not the distance of an individual normal state. Spectral dispersion explains the pure/mixed difference. Along a sequence becoming maximally distant, let the decreasing eigenvalues converge to ℓj\ell_jℓj​, and set θ=1−∑jℓj\theta=1-\sum_j\ell_jθ=1−∑j​ℓj​. This is spectral mass spread among vanishing eigenvalues, not probability discarded from the source. With independent mean-one exponential variables EjE_jEj​, the sharp limiting moment capacity is Ap(ℓ)=p−2pE ⁣(∑jℓjEj+pθ)p.\mathcal A_p(\ell)=p^{-2p}\mathbb E\!\left(\sum_j\ell_jE_j+p\theta\right)^p.Ap​(ℓ)=p−2pE(∑j​ℓj​Ej​+pθ)p. Every limiting spectrum has a physical attaining sequence. At second order, A2=[(1+θ)2+∑jℓj2]/16\mathcal A_2=[(1+\theta)^2+\sum_j\ell_j^2]/16A2​=[(1+θ)2+∑j​ℓj2​]/16, which gives the exact minimum dispersed mass θmin⁡(s)=8e−s−1,log⁡4≤s≤log⁡8.\theta_{\min}(s)=\sqrt{8e^{-s}-1},\qquad \log4\le s\le\log8.θmin​(s)=8e−s−1​,log4≤s≤log8. Quantitative refinements. The article retains explicit bounds and one-photon probability rigidity for the stated number-diagonal and coherent-label source classes, minimum limiting photon costs of one for mixed sources and two for pure sources, and the sharp number-diagonal distance slope 4/e4/\sqrt e4/e​. These finer source-class conclusions are kept distinct from the universal plateau theorem. At second order, the moment is also half-loss output purity and a two-copy interference observable. Version 2.0 extends the earlier source-class analysis to the full-source earliest platforms and limiting-spectrum capacity. The integrated 28-page PDF includes complete proofs and three figure groups. It determines the plateau onsets and spectral capacities.
How much distinguishability can a noisy quantum channel preserve when its input is correlated with an untouched quantum memory? This paper gives an exact finite semidefinite program for that question for every finite-dimensional completely positive, trace-preserving channel. The complete coefficient compares joint states with the same memory marginal. The conditional coefficient instead subtracts the distinguishability already present in the memory. Both coefficients, and conditional mutual-information contraction with a separate conditioning system, equal one minus the reciprocal of an optimization in the channel’s unnormalized Choi matrix. The equivalent inverse formula retains an essential range condition, including for singular channels. The proof solves a centered matrix-Gram problem, shows that its optimization center can be positive, and constructs a signed observable lift through an erasure channel. Two weighted quadratic inequalities give the universal entropy bound. A rare coherent memory sector proves sharpness while preserving both local marginals exactly. A memory of bounded dimension suffices, including with prescribed faithful input marginals. The same program gives exact coefficients for regular modular Petz divergences whose generators have finite endpoint value and recession slope, every fixed Petz Rényi order strictly between zero and one, and quantum Chernoff divergence. Further results include a harmonic dual, an exact tensor-product law, a strict-contraction criterion, a sharp comparison with positive quantum Doeblin minorization, and explicit classical, depolarizing and orthogonal-unitary-noise formulas. Ordinary quantum mutual-information contraction is a different optimization. For qubit depolarization the paper determines its exact value on the full completely positive parameter interval. At shrinkage one half, ordinary relative-entropy, ordinary mutual-information and complete entropy contraction are respectively 1/4, 1/3 and 3/8. In higher dimension, bounds and a strict separation are proved without asserting an exact ordinary-MI formula. A discontinuity between nearly complete amplitude damping and pure reset rules out uniformly bounded-use estimation of the exact coefficient from an unknown channel, even with adaptive experiments and quantum ancillas. Thus evaluation from exact supplied Choi data does not imply stable experimental acquisition. Rare sharpness witnesses can carry vanishing absolute information; the results do not assert a communication rate, physical simulator, general recovery map or polynomial cost in physical qubit count.
How much high-photon quantum coherence is controlled by a nonnegative phase-space distribution? This article gives quantitative guarantees for approximating an optical state by a finite photon-number block, and constructs physical states showing why the guarantee cannot be inferred from mean energy or Wigner positivity alone at the strength available for a Gaussian-mixture source. Source classes and scope. The comparison concerns one optical mode with [Q,P]=i[Q,P]=i[Q,P]=i, mean photon number at most E>0E>0E>0, and a known parity centre. The full class WE\mathcal W_EWE​ consists of normal Wigner-positive states at the exact vacuum min-entropy endpoint, equivalently states supported on the even photon sector in that frame. The subclass GE\mathcal G_EGE​ consists of actual probability mixtures of centred pure squeezed vacua. Mixture labels are unavailable, component energies may be arbitrarily large, and their true average energy is charged. Infinite photon tails are retained. A hierarchy of coherent-tail guarantees. Let PLP_LPL​ project onto photon numbers at most LLL, and define the cutoff error by eL(ρ)=12∥ρ−PLρPL∥1.e_L(\rho)=\tfrac12\|\rho-P_L\rho P_L\|_1.eL​(ρ)=21​∥ρ−PL​ρPL​∥1​. This quantity includes both discarded photon population and coherence across the cutoff. A mean-energy constraint alone gives the standard OE(L−1/2)O_E(L^{-1/2})OE​(L−1/2) guarantee. At the exact Wigner endpoint, the article proves sup⁡ρ∈WEeL(ρ)≤CE(L+1)−3/4,sup⁡ρ∈GEeL(ρ)≤4EL+1,L≥2.\sup_{\rho\in\mathcal W_E}e_L(\rho)\le C_E(L+1)^{-3/4},\qquad \sup_{\rho\in\mathcal G_E}e_L(\rho)\le \frac{4E}{L+1},\qquad L\ge2.supρ∈WE​​eL​(ρ)≤CE​(L+1)−3/4,supρ∈GE​​eL​(ρ)≤L+14E​,L≥2. The stronger endpoint bound combines operator positivity and Wigner positivity on the same density kernel, followed by a controlled conversion to trace norm. It requires only mean energy, without a uniform second photon moment. Physical coherence beyond the Gaussian-mixture bound. A family τd∈WE\tau_d\in\mathcal W_Eτd​∈WE​ has exact energy EEE, covariance (E+1/2)I(E+1/2)I(E+1/2)I, strictly positive Wigner functions and nonzero population at every even photon number. For cutoffs Ld=Θ(d3)L_d=\Theta(d^3)Ld​=Θ(d3), eLd(τd)≥cEd−5/2,(Ld+1)eLd(τd)⟶∞.e_{L_d}(\tau_d)\ge c_Ed^{-5/2},\qquad (L_d+1)e_{L_d}(\tau_d)\longrightarrow\infty.eLd​​(τd​)≥cE​d−5/2,(Ld​+1)eLd​​(τd​)⟶∞. Thus the Gaussian-mixture L−1L^{-1}L−1 law fails uniformly on the full endpoint class. The construction preserves coherence through an actual optical partial trace and pays its physical energy cost. Positive phase mixing further yields distinct states with identical complete photon statistics and covariance but nearly indistinguishable bounded optical responses. The rare bright branches obey the prescribed mean-energy budget; no efficient experimental preparation protocol or uniform branch-energy bound is assumed. Consequences for a specified measurement system. Each independent copy is measured by bounded centred squeezing followed by a vacuum/nonvacuum binary test. At fixed energy, squeezing range and confidence, let nX(δ)n_X(\delta)nX​(δ) be the minimum worst-case number of copies for trace-distance accuracy δ\deltaδ over source class XXX. As δ↓0\delta\downarrow0δ↓0, the bounds are log⁡nGE(δ)=Θ(δ−1),cδ−6/5≤log⁡nWE(δ)≤Cδ−4/3.\log n_{\mathcal G_E}(\delta)=\Theta(\delta^{-1}),\qquad c\delta^{-6/5}\le\log n_{\mathcal W_E}(\delta)\le C\delta^{-4/3}.lognGE​​(δ)=Θ(δ−1),cδ−6/5≤lognWE​​(δ)≤Cδ−4/3. These are measurement-restricted reconstruction costs. The lower bounds allow adaptive randomized settings. The upper bounds use finite designs and estimates in the original source class, with the energy budget respected for every record; they establish statistical existence rather than an efficient catalogue-generation algorithm. The full-class exponents remain distinct lower and upper bounds. The article also characterizes infinite-order Wigner entropy equality by displaced even-parity support and corrects the pure-Gaussian-only equality clause at infinite order in Dias and Prata (2023). The 27-page integrated article includes the complete proofs, figures and references.