Pith. sign in

REVIEW 4 minor 35 references

Coupling Does Not Reduce the Auxiliary-Mode Count for $1/|\omega|$ Spectra in Passive Lindblad Networks

T0 review · 0 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read For passive Gaussian Lindblad networks, coherent intermode coupling cannot beat N independent damped modes at zero detuning when fitting a 1/|ω| spectrum: the optimal error is the degree-2N Zolotarev error.

desk verdict Exact no-advantage theorem for 1/|ω| spectra: the proof holds up, with only a citation-level soft spot in the Zolotarev lemma. read the letter →

arxiv 2608.04539 v1 pith:SWKDWPF5 submitted 2026-08-05 quant-ph

Qing-Ao Xiang , Yan Liu , Xin-Yuan Yang , Ya-Ju Song This is my paper · ORCID
classification quant-ph
keywords openquantumsystemsnon-MarkovianbathspseudomodesLindbladnetworks1/fnoiseZolotarevapproximationrationalspectroscopy
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper asks whether, at a fixed number of auxiliary modes N, adding coherent intermode coupling, collective dissipation, or nonnormal structure can reduce the best possible maximum relative error when fitting a 1/|ω| spectrum over a finite two-sided frequency band. It proves the answer is no: the optimal error of the general coupled class equals the optimal error of the much simpler uncoupled diagonal subclass, and both equal the degree-2N Zolotarev error for rational approximation of the sign function. The optimum is achieved by N independent damped auxiliary modes at zero detuning, i.e., by a sum of Lorentzians. As a result, the O(N²) parameters available to a coupled network do not reduce the number of modes needed for a prescribed tolerance; they only move poles and residues within a fixed rational-degree constraint. This exact relation yields closed-form answers for the minimum mode count for a prescribed dynamic range and tolerance, and for the maximum dynamic range attainable at a fixed mode budget.

What carries the argument

The load-bearing object is the rational-degree bound: an $n$-mode passive Gaussian Lindblad network yields a spectrum $J(\omega)$ whose numerator and denominator degrees are at most $2n-2$ and $2n$ after cancellation, so $\omega J(\omega)$ has rational type at most $(2n-1,2n)$, regardless of whether the drift matrix $M=\Gamma+iH$ is normal or diagonalizable. This converts spectral fitting into the fourth Zolotarev problem of approximating $\operatorname{sgn}(\omega)$ on two disjoint intervals. The construction uses Zolotarev's odd optimal rational approximant of even degree $2N$; after factoring out $\omega$, the remaining rational function of $\omega^2$ lies exactly in the class $\mathcal{RR}_{N-1,N}([1,R^2])$ for approximating $x^{-1/2}$. The crucial structural fact imported from Zolotarev theory is that this approximant can be chosen with all partial-fraction residues positive, which is what makes the realization by independent damped modes possible. The equality then follows by sandwiching: Zolotarev error is a lower bound for the coupled class, the diagonal subclass attains it, and the coupled class cannot do worse than the diagonal subclass.

What would settle it

Take a small instance such as R=4, N=2 and globally search the full general coupled class of stable passive Lindblad networks for a maximum relative error over $[-4,-1]\cup[1,4]$ strictly below $e_4(1/4)$; finding such a network would disprove the equality. A complementary check is to verify the imported positive-residue property for the known Zolotarev approximants: if some degree-2N optimal sign approximant necessarily has a negative residue, the physical realization step would fail.

Watch

Extended reading notes

Core claim

The central claim is the equality $E^{\mathrm{coup}}_N(R)=E^{\mathrm{diag}}_N(R)=e_{2N}(R^{-1})$, where $E^{\mathrm{coup}}_N(R)$ and $E^{\mathrm{diag}}_N(R)$ are the optimal maximum relative errors over $\Omega_R=[-R,-1]\cup[1,R]$ for the general coupled class and its uncoupled diagonal subclass, and $e_{2N}(R^{-1})$ is the optimal error of the degree-$2N$ fourth Zolotarev problem for approximating $\operatorname{sgn}(\omega)$ on $[-1,-1/R]\cup[1/R,1]$. The proof has three legs: a rational-degree bound showing that any stable passive $n$-mode network produces $\omega J(\omega)$ of type at most $(2n-1,2n)$; an isometry showing that sign approximation on the two-sided band is minimax equivalent to relative-error approximation of $x^{-1/2}$ on $[1,R^2]$; and a realization argument showing that the optimal rational function has an all-positive partial-fraction expansion, so it is exactly a sum of Lorentzians generated by $N$ independent damped modes at zero detuning. Because the optimum lies inside the uncoupled diagonal subclass, the full coupled class cannot beat it.

Load-bearing premise

The proof's load-bearing assumption is an imported result from classical approximation theory: the optimal rational sign approximant can be chosen so that, when expanded in simple fractions, every coefficient is positive and every pole lies on the negative real axis. This is what turns a mathematical optimum into a physical array of damped modes; the paper cites this structural fact but does not prove it here.

Editorial extensions

If this is right

  • No passive coupled Lindblad network with N modes can attain a maximum relative error below the Zolotarev error $e_{2N}(R^{-1})$; the rational-degree bound is the only constraint that matters for this benchmark.
  • The minimum number of auxiliary modes needed for a prescribed tolerance and dynamic range is given by an exact inversion of the error relation, with a large-range, small-tolerance asymptotic of the form $\ln(4R)\ln(4/\varepsilon)/\pi^2$.
  • The maximum positive-frequency dynamic range attainable with a fixed mode budget and tolerance follows by the same inversion; for 1%, 5%, and 10% tolerances each added mode asymptotically buys about 0.72, 0.98, and 1.16 decades of span.
  • The optimum is physically simple: N independent damped modes at zero detuning suffice, so one need not engineer coherent intermode coupling or collective dissipation to reach the fundamental limit.
  • Any candidate network's optimality can be certified by checking equioscillation of $\omega J(\omega)$ at the $2N+1$ alternation points, as done numerically in the paper.

Reading between the lines

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

  • Editorial extension: the no-advantage result relies on the combination of a single Hermitian bath operator and the exact $1/|\omega|$ target; for multi-channel baths, asymmetric detailed-balance spectra, or a white-noise feedthrough term, the rational-degree bound changes and coherent coupling may plausibly help. This is not claimed in the paper.
  • Editorial extension: the exact mode-count formula provides a practical lower-bound test for any pseudomode or environment model: if a measured $1/|\omega|$ noise band is claimed to be reproduced with fewer modes than the paper's $N_{\min}$, the model either lies outside this passive class or the claim should be re-examined.
  • Editorial extension: one could apply the same rational-degree-plus-realization strategy to targets like $1/|\omega|^{\alpha}$ with $\alpha\neq 1$; the sign-function isometry would be replaced by a relative-error problem on one interval, and it is an open question whether positive-residue realizations persist for all such exponents.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

0 major / 4 minor

Summary. The paper studies whether coherent intermode coupling can reduce the number of auxiliary modes needed to approximate a 1/|ω| spectrum over a finite two-sided frequency band in passive, number-conserving Gaussian Lindblad networks. The central result, Eq. (12), states that for any mode budget N and dynamic range R>1, the optimal maximum relative error for the general coupled class equals that of the uncoupled diagonal subclass, and both equal the degree-2N Zolotarev error e_{2N}(R^{-1}) for sign approximation on two intervals. The authors prove the lower bound via a rational-degree constraint imposed by the mode count, construct an explicit upper-bound realization using N independent damped auxiliary modes at zero detuning, and then derive closed-form inversions for the minimum mode count and maximum dynamic range in terms of elliptic integrals. The paper includes a supplemental material with detailed proofs of the rational-degree bound, the sign/inverse-square-root isometry, the positive-residue partial-fraction construction, and extensive numerical cross-checks.

Significance. If the central theorem holds, it settles a structural question in pseudomode and non-Markovian bath modeling: despite having O(N^2) real parameters, coupled passive Lindblad networks offer no advantage over independent damped modes for this canonical 1/|ω| benchmark. The paper's strengths include a clean reduction of the physical problem to a classical Zolotarev problem, an explicit construction with positive residues that is directly realizable as a CPTP Lindblad network, and exact elliptic-function formulas for mode count and dynamic range. The numerical verification is unusually thorough: the authors report 70-digit agreement between the Jacobi product formula and the modular equation in 35 cases, exact agreement of the closed-form mode-count formula with sequential search in all 84 tested cases, and agreement with direct nonconvex optimization to better than 10^{-4}. The derivation is parameter-free in the sense that no parameters are fitted to the target spectrum; all quantities are determined by the band edges and tolerance.

minor comments (4)
  1. [Supplemental Sec. S3, Lemma S3] The upper-bound chain in Eq. (15) relies entirely on Lemma S3, which asserts that the degree-2N Zolotarev problem admits an odd extremal of type (2N-1,2N) and that after factoring out the argument the remaining rational function has the required type. This lemma is quoted from the arXiv v1 version of Ref. [27] rather than proved in the paper. Because this is the only load-bearing step not derived self-containedly, I recommend that the authors either supply a proof in the Supplemental Material or state the precise theorem with equation numbers from the published version of Ref. [27]. The 70-digit numerical checks and agreement with direct optimization make a genuine error unlikely, but the citation should be made verifiable.
  2. [Supplemental Sec. S7] In the direct-optimization cross-check, the text says that 'a linear program computes the minimax error level on the current exchange set' but does not give the variables or constraints of that linear program. Please add a one-sentence description of the LP (e.g., minimize the maximal deviation at the exchange points subject to the rational ansatz) so the procedure is reproducible.
  3. [Supplemental Eq. (S59)] The notation δα(R) should be typeset as δ_α(R) to avoid confusion with a Kronecker delta, and the subscript α should be defined explicitly as the power-law exponent.
  4. [Main text, Eq. (13) and surrounding text] The sentence 'Returning to the frequency variable ω then defines the corresponding spectrum' could be clearer: the spectrum is defined by J_opt(ω)=r_opt(ω^2), which is then identified with a sum of Lorentzians. Please rephrase to avoid the appearance of a circular definition.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the main equality follows from a degree-relaxation lower bound plus an explicit positive-residue Zolotarev construction, with the only external input a cited classical lemma that the paper independently verifies.

full rationale

The derivation chain is self-contained apart from standard external Zolotarev theory. The lower bound is a genuine relaxation: every physical network produces q(ω)=ωJ(ω) in RR_{2N−1,2N}(Ω_R) ⊂ RR_{2N,2N}(Ω_R) via Lemma S1, which is proved from the adjugate formula for (M−iωI_n)^{-1} without assuming normality or diagonalizability. Taking the infimum over the larger rational class gives e_{2N}(R^{-1}) ≤ E_coup_N(R); no parameter is fitted to the target spectrum. The upper bound constructs an explicit uncoupled Lorentzian spectrum from the Jacobi product form of the even-degree Zolotarev approximant, with the odd type-(2N−1,2N) structure cited from Ref. [27] and stated in explicit product form in Eq. (S19), and then proves all partial-fraction residues positive from pole–zero interlacing in Lemma S4 and Eq. (S32). The physical parameters γ_j and |g_j|^2 are assigned algebraically from the residues via Eq. (S33), not fitted. The mode-count and dynamic-range formulas, Eqs. (19)–(21), are exact inversions of the theorem using the strictly monotone function μ(k), and the asymptotic forms follow from standard complete-elliptic-integral expansions. The only load-bearing external input is the classical Zolotarev parity/type fact cited to Ref. [27]; that citation is not self-citational, is reproduced in explicit product form, and is independently cross-checked in Sec. S7 to 70 significant digits in 35 cases plus direct nonconvex optimization in 9 cases. No equation is defined in terms of the quantity it is used to predict, and no fitted input is relabeled as a prediction. Therefore the paper exhibits no significant circularity.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The proof uses standard Zolotarev rational approximation theory and standard quantum regression/CPTP Lindblad machinery. No free parameters are fitted; the only inputs are the mode budget N, dynamic range R, and tolerance ε, which are variables of the problem, not fitted constants. No new entities are introduced.

assumptions (4)
  • standard math Zolotarev theory: the minimax error for sign approximation on two intervals is attained by an odd rational function of type (2N-1, 2N) with the given product form
    Cited from Ref. [27] Sec. 3.1, used in Lemma S3 to establish the upper bound and the positive-residue construction.
  • domain assumption Quantum regression theorem: bath correlation function follows homogeneous evolution e^{-Mt} in the stationary vacuum state
    Used in Eq. (3) to express the spectrum as g†(M-iωI)^{-1}g. Standard in open quantum systems; cited to Lax [29].
  • domain assumption The system couples to the auxiliary network through a single Hermitian bath operator B = g†b + b†g, with no feedthrough term
    Scope condition of the model class; stated in main text after Eq. (2). The result does not apply if this is violated.
  • domain assumption The auxiliary network is passive, number-conserving Gaussian, and in a stationary vacuum state
    Defines the model class; stated in the model section.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Coupling Does Not Reduce the Auxiliary-Mode Count for $1/|\omega|$ Spectra in Passive Lindblad Networks." pith.science (2026). https://pith.science/paper/SWKDWPF5

@misc{pith2026260804539,
  author       = {Pith},
  title        = {Pith review of: Coupling Does Not Reduce the Auxiliary-Mode Count for $1/|\omega|$ Spectra in Passive Lindblad Networks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SWKDWPF5}},
  note         = {Machine review of arXiv:2608.04539}
}
abstract

Representing continuous environments by finitely many Markovian auxiliary modes is fundamental in non-Markovian open quantum systems, yet a critical question remains: at a fixed mode budget, can coherent intermode coupling reduce the spectral approximation error? We prove that intermode coupling offers no advantage when passive, number-conserving Gaussian Lindblad auxiliary networks approximate a $1/|\omega|$ spectrum over a finite two-sided frequency band. For any mode budget $N$, the general coupled class and its uncoupled diagonal subclass share the same optimal error, which is exactly the degree-$2N$ Zolotarev error for sign approximation. This optimum is attainable by $N$ independent damped auxiliary modes at zero detuning. The result holds when the auxiliary network is in a stationary vacuum state, the system couples to it via a single Hermitian bath operator, and no white-noise feedthrough term is present. Consequently, although a general coupled network has $O(N^{2})$ real parameters, coherent intermode coupling, collective dissipation, and nonnormal structure cannot reduce the number of auxiliary modes required to reach a prescribed tolerance. This exact relation yields both the minimum mode count for a prescribed positive-frequency dynamic range and tolerance, and the maximum dynamic range attainable for a prescribed mode budget and tolerance.

Figures

Figures reproduced from arXiv: 2608.04539 by the authors.

Figure 1
Figure 1. FIG. 1 [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

35 extracted references · 30 canonical work pages

  1. [27]

    E. S. Gawlik, SIAM J. Matrix Anal. Appl.40, 696 (2019)

    work page 2019

  2. [1]

    A. D. Baczewski and S. D. Bond, J. Chem. Phys.139, 044107 (2013)

    work page 2013

  3. [2]

    Gorini, A

    V. Gorini, A. Kossakowski, and E. C. G. Sudarshan, J. Math. Phys.17, 821 (1976)

    1976

  4. [3]

    Lindblad, Commun

    G. Lindblad, Commun. Math. Phys.48, 119 (1976)

    1976

  5. [4]

    B. M. Garraway, Phys. Rev. A55, 2290 (1997)

    1997

  6. [5]

    Mazzola, S

    L. Mazzola, S. Maniscalco, J. Piilo, K.-A. Suominen, and B. M. Garraway, Phys. Rev. A80, 012104 (2009)

    work page 2009

  7. [6]

    Tamascelli, A

    D. Tamascelli, A. Smirne, S. F. Huelga, and M. B. Plenio, Phys. Rev. Lett.120, 030402 (2018)

    work page 2018

  8. [7]

    Pleasance, B

    G. Pleasance, B. M. Garraway, and F. Petruccione, Phys. Rev. Research2, 043058 (2020)

    2020

Show all 35 references
  1. [8]

    Trivedi, D

    R. Trivedi, D. Malz, and J. I. Cirac, Phys. Rev. Lett.127, 250404 (2021)

    2021

  2. [9]

    Mascherpa, A

    F. Mascherpa, A. Smirne, A. D. Somoza, P. Fern´ andez- Acebal, S. Donadi, D. Tamascelli, S. F. Huelga, and M. B. Plenio, Phys. Rev. A101, 052108 (2020)

    2020

  3. [10]

    Dorda, M

    A. Dorda, M. Nuss, W. von der Linden, and E. Arrigoni, Phys. Rev. B89, 165105 (2014)

    2014

  4. [11]

    Dorda, M

    A. Dorda, M. Ganahl, H. G. Evertz, W. von der Linden, and E. Arrigoni, Phys. Rev. B92, 125145 (2015)

    2015

  5. [12]

    Lednev, F

    M. Lednev, F. J. Garc ´ ıa-Vidal, and J. Feist, Phys. Rev. Lett.132, 106902 (2024)

    2024

  6. [13]

    Huang, G

    Z. Huang, G. Park, G. K.-L. Chan, and L. Lin, Phys. Rev. Lett.136, 090403 (2026)

    2026

  7. [14]

    Alford, L

    W. Alford, L. P. Bettmann, and G. T. Landi, Phys. Rev. B113, 205143 (2026). 6

    2026

  8. [15]

    G. Park, Z. Huang, Y. Zhu, C. Yang, G. K.-L. Chan, and L. Lin, Phys. Rev. B110, 195148 (2024)

    2024

  9. [16]

    Thoenniss, I

    J. Thoenniss, I. Vilkoviskiy, and D. A. Abanin, Phys. Rev. B112, 155114 (2025)

    2025

  10. [17]

    Vilkoviskiy and D

    I. Vilkoviskiy and D. A. Abanin, Phys. Rev. B109, 205126 (2024)

    2024

  11. [18]

    Huang, Z

    Z. Huang, Z. Ding, K. Wang, J. Kaye, X. Li, and L. Lin, Provably efficient long-time exponential decompositions of Non-Markovian Gaussian baths (2026), 2603.25708

    2026 arXiv

  12. [19]

    Yoshihara, K

    F. Yoshihara, K. Harrabi, A. O. Niskanen, Y. Nakamura, and J.-S. Tsai, Phys. Rev. Lett.97, 167001 (2006)

    2006

  13. [20]

    Bylander, S

    J. Bylander, S. Gustavsson, F. Yan, F. Yoshihara, K. Harrabi, G. Fitch, D. G. Cory, Y. Nakamura, J.-S. Tsai, and W. D. Oliver, Nat. Phys.7, 565 (2011)

    2011

  14. [21]

    Machlup, J

    S. Machlup, J. Appl. Phys.25, 341 (1954)

    1954

  15. [22]

    Sauv´ e and G

    R. Sauv´ e and G. Szabo, J. Theor. Biol.113, 501 (1985)

    1985

  16. [23]

    Pellegrini, B

    B. Pellegrini, B. Neri, and R. Saletti, Alta Frequenza55, 245 (1986)

    1986

  17. [24]

    Milotti, Phys

    E. Milotti, Phys. Rev. E51, 3087 (1995)

    1995

  18. [25]

    N. I. Akhiezer,Theory of Approximation(Dover, New York, 1992)

    1992

  19. [26]

    Istace and J.-P

    M.-P. Istace and J.-P. Thiran, SIAM J. Numer. Anal.32, 249 (1995)

    1995

  20. [28]

    L. N. Trefethen and H. D. Wilber, SIAM J. Sci. Comput. 47, A2205 (2025)

    2025

  21. [29]

    Lax, Phys

    M. Lax, Phys. Rev.129, 2342 (1963)

    1963

  22. [30]

    Cywi´ nski, R

    L. Cywi´ nski, R. M. Lutchyn, C. P. Nave, and S. Das Sarma, Phys. Rev. B77, 174509 (2008)

    2008

  23. [31]

    G. A. ´Alvarez and D. Suter, Phys. Rev. Lett.107, 230501 (2011)

    2011

  24. [32]

    Cerfontaine, T

    P. Cerfontaine, T. Hangleiter, and H. Bluhm, Phys. Rev. Lett.127, 170403 (2021)

    2021

  25. [33]

    L. B. Vieira, S. Milz, G. Vitagliano, and C. Budroni, Quantum8, 1224 (2024)

    2024

  26. [34]

    Storn and K

    R. Storn and K. Price, J. Global Optim.11, 341 (1997)

    1997

  27. [35]

    J. A. Nelder and R. Mead, Comput. J.7, 308 (1965). 7 Supplemental Material Coupling Does Not Reduce the Auxiliary-Mode Count for1/|ω|Spectra in Passive Lindblad Networks S1. PHYSICAL MODEL, SPECTRAL POSITIVITY, AND RESTORA TION OF PHYSICAL UNITS The proof in this Supplemental ...

    1965

Pith tools

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