pith. sign in

arxiv: 2606.09791 · v1 · pith:4XDBUHU3new · submitted 2026-06-08 · ✦ hep-lat · cond-mat.str-el· hep-th

Certified spectral functions from lattice Monte Carlo data

Sophie Mutzel , Antoine Tilloy This is my paper

Pith reviewed 2026-06-27 14:00 UTC · model grok-4.3

classification ✦ hep-lat cond-mat.str-elhep-th
keywords spectral densitylattice Monte Carlosemidefinite programmingreflection positivityinverse problemEuclidean correlation functionsconvex optimizationphi^4 theory
0
0 comments X

The pith

Spectral density bounds from lattice Monte Carlo data are obtained by relaxing the dual problem to a hierarchy of semidefinite programs.

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

Instead of reconstructing a single spectral density from Euclidean correlation functions measured in lattice Monte Carlo simulations, the paper bounds any linear functional of that density. The bounds incorporate both the Monte Carlo data with its error bars and the reflection positivity constraint supplied by the lattice action. This produces an infinite-dimensional convex optimization problem whose dual is then relaxed to a sequence of finite semidefinite programs. The resulting numerical bounds remain valid at every finite relaxation order and converge rapidly until the remaining uncertainty is set by the Monte Carlo statistics alone. The same procedure also returns an infeasibility certificate when the supplied data or error estimates are internally inconsistent.

Core claim

Given Monte Carlo estimates of Euclidean correlators and the reflection positivity of the lattice action, the possible values of linear functionals of the spectral density are constrained by a convex set. The dual of this problem can be relaxed to a sequence of finite-dimensional semidefinite programs whose solutions provide rigorous upper and lower bounds that converge to the true optimum as the relaxation order increases.

What carries the argument

The hierarchy of finite semidefinite program relaxations of the dual convex optimization problem over admissible spectral densities constrained by reflection positivity.

If this is right

  • Any linear functional of the spectral density can be bounded rigorously with off-the-shelf semidefinite programming solvers.
  • The methodological error in the bounds decreases until it is smaller than the Monte Carlo statistical error.
  • An infeasible semidefinite program certifies that the input Monte Carlo data or its quoted uncertainties are implausible.
  • The procedure applies to any lattice theory whose action satisfies reflection positivity.
  • Low-order relaxations already reach the Monte Carlo-limited regime for the two-dimensional phi^4 theory example.

Where Pith is reading between the lines

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

  • The same relaxation strategy could be applied to other lattice inverse problems that admit a convex formulation with positivity constraints.
  • Combining the bounds with additional lattice observables such as masses or matrix elements would produce tighter certified intervals for physical quantities.
  • The method provides a quantitative test for whether Monte Carlo error bars have been underestimated before any spectral reconstruction is attempted.
  • Application to higher-dimensional or gauge theories would require only that reflection positivity holds and that the requisite correlation functions can be measured.

Load-bearing premise

The lattice action must satisfy reflection positivity to supply the positivity constraint on admissible spectral densities.

What would settle it

Generate Monte Carlo data for a model whose exact spectral density is known analytically, then verify that the computed bounds always contain the true value of each tested functional and tighten until limited only by the Monte Carlo error bars.

Figures

Figures reproduced from arXiv: 2606.09791 by Antoine Tilloy, Sophie Mutzel.

Figure 1
Figure 1. Figure 1: FIG. 1 [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. Rigorous piecewise constant (turquoise) and piece ± [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4. Rigorous upper and lower bounds on the smeared [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗
read the original abstract

The Monte Carlo method, applied to lattice quantum field theory, gives access to Euclidean correlation functions with well-understood error bars. Recovering the observables one cares about, such as the spectral density, requires solving an ill-posed inverse problem, usually tackled with heuristics that lose rigorous control of the error. Instead of trying to find the ``best'' spectral density $\rho(\omega)$, we ask how small or large linear functionals $\int_{\mathbb{R}^+} G(\omega) \rho(\omega) \mathrm{d} \omega$ of it can be, given the Monte Carlo data and the reflection positivity of the lattice action. This is a convex but infinite-dimensional problem. We show how its dual can be rigorously relaxed into a hierarchy of finite semidefinite programs, solvable with standard solvers and enjoying strong convergence guarantees. The resulting bounds are rigorous even when the relaxation is not tight, and converge quickly to the regime where the error is entirely dominated by Monte Carlo statistics. The method also flags implausible Monte Carlo data, for instance underestimated error bars, through an infeasibility certificate. We demonstrate it on lattice $\phi^4$ theory in two dimensions.

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.

Referee Report

0 major / 3 minor

Summary. The manuscript formulates bounding linear functionals of the spectral density as an infinite-dimensional convex optimization problem constrained by Monte Carlo data intervals and reflection-positivity moment matrices from the lattice transfer matrix. It relaxes the dual to a Lasserre-type hierarchy of finite SDPs, invokes existing convergence results under the archimedean condition supplied by exponential decay of correlators, and shows that the resulting bounds remain rigorous (even when the relaxation is not tight) and rapidly become limited only by Monte Carlo statistics. The approach is demonstrated on two-dimensional φ⁴ theory and includes an infeasibility certificate for inconsistent input data.

Significance. If the claimed relaxation and convergence hold, the work supplies a rigorous, non-heuristic route to certified bounds on spectral observables that is directly usable with existing SDP solvers. Explicit credit is due for the parameter-free character of the bounds once the hierarchy is solved, the built-in consistency check for Monte Carlo error bars, and the demonstration that the method reaches the Monte Carlo-dominated regime at modest hierarchy depth. This is a substantive contribution to the treatment of ill-posed inverse problems in lattice field theory.

minor comments (3)
  1. [§4] The statement that convergence follows from “existing results on the hierarchy under the archimedean condition” (abstract and §4) would benefit from an explicit citation to the precise theorem invoked and a one-sentence verification that the exponential decay of the lattice correlators indeed supplies the required archimedean quadratic module.
  2. [§5] In the numerical demonstration (§5), the tables or figures reporting the SDP bounds should state the hierarchy level (k) at which the reported intervals stabilize to within Monte Carlo error; without this, it is difficult to judge how quickly the “quick convergence” asserted in the abstract is realized.
  3. [§3] Notation for the moment matrices constructed from the lattice action (Eqs. (12)–(15)) is clear, but a short appendix listing the explicit block structure for the lowest two levels of the hierarchy would help readers reproduce the SDP constraints.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the positive assessment of the manuscript, the recognition of its contributions to rigorous bounds on spectral observables, and the recommendation for minor revision. No specific major comments were provided in the report.

Circularity Check

0 steps flagged

No significant circularity identified

full rationale

The paper's central derivation applies the standard Lasserre moment-SOS hierarchy to the dual of a convex problem whose constraints are supplied externally by reflection positivity of the lattice action (encoded via moment matrices from the transfer matrix) and by interval bounds on Monte Carlo data. Convergence guarantees are invoked from existing results on the hierarchy under the archimedean condition, which is independent of the present work. No step reduces by construction to a fitted parameter, self-citation chain, or self-definitional ansatz; the SDP relaxations and infeasibility certificates are standard and externally verifiable.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the domain assumption of reflection positivity and the validity of the SDP relaxation hierarchy; no free parameters or invented entities are mentioned in the abstract.

axioms (1)
  • domain assumption The lattice action satisfies reflection positivity
    This property constrains the admissible spectral densities in the optimization problem.

pith-pipeline@v0.9.1-grok · 5730 in / 1152 out tokens · 25908 ms · 2026-06-27T14:00:01.911711+00:00 · methodology

discussion (0)

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

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Kernel transformations and bounds for smeared spectral functions

    hep-lat 2026-06 unverdicted novelty 5.0

    Framework for exact and approximate kernel transformations between smeared spectral functions, including systematic error bounds computable from input data.

Reference graph

Works this paper leans on

52 extracted references · 1 canonical work pages · cited by 1 Pith paper

  1. [1]

    Durret al.(BMW), Ab-Initio Determination of Light Hadron Masses, Science322, 1224 (2008), arXiv:0906.3599 [hep-lat]

    S. Durret al.(BMW), Ab-Initio Determination of Light Hadron Masses, Science322, 1224 (2008), arXiv:0906.3599 [hep-lat]

    Pith/arXiv arXiv 2008

  2. [2]

    Aokiet al.(Flavour Lattice Averaging Group (FLAG)), FLAG review 2024, Phys

    Y. Aokiet al.(Flavour Lattice Averaging Group (FLAG)), FLAG review 2024, Phys. Rev. D113, 014508 (2026), arXiv:2411.04268 [hep-lat]

    Pith/arXiv arXiv 2024

  3. [3]

    Boccalettiet al., Hybrid calculation of hadronic vac- uum polarization in muon g−2 to 0.48%, Nature653, 373 (2026), arXiv:2407.10913 [hep-lat]

    A. Boccalettiet al., Hybrid calculation of hadronic vac- uum polarization in muon g−2 to 0.48%, Nature653, 373 (2026), arXiv:2407.10913 [hep-lat]

    Pith/arXiv arXiv 2026

  4. [4]

    Jarrell and J

    M. Jarrell and J. Gubernatis, Bayesian inference and the analytic continuation of imaginary-time quantum monte carlo data, Physics Reports269, 133 (1996)

    1996

  5. [5]

    Asakawa, T

    M. Asakawa, T. Hatsuda, and Y. Nakahara, Maxi- mum entropy analysis of the spectral functions in lattice QCD, Prog. Part. Nucl. Phys.46, 459 (2001), arXiv:hep- lat/0011040

    arXiv 2001

  6. [6]

    Burnier and A

    Y. Burnier and A. Rothkopf, Bayesian approach to spec- tral function reconstruction for euclidean quantum field theories, Phys. Rev. Lett.111, 182003 (2013)

    2013

  7. [7]

    K. S. D. Beach, Identifying the maximum entropy method as a special limit of stochastic analytic continu- ation (2004), arXiv:cond-mat/0403055 [cond-mat.str-el]

    Pith/arXiv arXiv 2004

  8. [8]

    Rothkopf, Bayesian inference of real-time dynam- ics from lattice QCD, Front

    A. Rothkopf, Bayesian inference of real-time dynam- ics from lattice QCD, Front. Phys.10, 1028995 (2022), arXiv:2208.13590 [hep-lat]

    arXiv 2022

  9. [9]

    Rothkopf, Bayesian inference of nonpositive spectral functions in quantum field theory, Phys

    A. Rothkopf, Bayesian inference of nonpositive spectral functions in quantum field theory, Phys. Rev. D95, 056016 (2017)

    2017

  10. [10]

    Backus and F

    G. Backus and F. Gilbert, Uniqueness in the inversion of inaccurate gross earth data, Philosophical Transactions of the Royal Society of London. Series A, Mathematical and Physical Sciences266, 123 (1970)

    1970

  11. [11]

    M. T. Hansen, H. B. Meyer, and D. Robaina, From deep inelastic scattering to heavy-flavor semileptonic decays: Total rates into multihadron final states from lattice QCD, Phys. Rev. D96, 094513 (2017), arXiv:1704.08993 [hep-lat]

    Pith/arXiv arXiv 2017

  12. [12]

    Hansen, A

    M. Hansen, A. Lupo, and N. Tantalo, Extraction of spec- tral densities from lattice correlators, Phys. Rev. D99, 094508 (2019), arXiv:1903.06476 [hep-lat]

    Pith/arXiv arXiv 2019

  13. [13]

    Tripolt, P

    R.-A. Tripolt, P. Gubler, M. Ulybyshev, and L. Von Smekal, Numerical analytic continuation of Euclidean data, Comput. Phys. Commun.237, 129 (2019), arXiv:1801.10348 [hep-ph]

    Pith/arXiv arXiv 2019

  14. [14]

    Schlessinger, Use of analyticity in the calculation of nonrelativistic scattering amplitudes, Phys

    L. Schlessinger, Use of analyticity in the calculation of nonrelativistic scattering amplitudes, Phys. Rev.167, 1411 (1968)

    1968

  15. [15]

    Bergamaschi, W

    T. Bergamaschi, W. I. Jay, and P. R. Oare, Hadronic structure, conformal maps, and analytic continuation, Phys. Rev. D108, 074516 (2023), arXiv:2305.16190 [hep- lat]. 10

    arXiv 2023

  16. [16]

    A. W. Sandvik, Stochastic method for analytic contin- uation of quantum monte carlo data, Phys. Rev. B57, 10287 (1998)

    1998

  17. [17]

    Buzzicotti, A

    M. Buzzicotti, A. De Santis, and N. Tantalo, Teaching to extract spectral densities from lattice correlators to a broad audience of learning-machines, Eur. Phys. J. C84, 32 (2024), arXiv:2307.00808 [hep-lat]

    arXiv 2024

  18. [18]

    L. Wang, S. Shi, and K. Zhou, Reconstructing spec- tral functions via automatic differentiation, Phys. Rev. D106, L051502 (2022), arXiv:2111.14760 [hep-ph]

    arXiv 2022

  19. [19]

    Fournier, L

    R. Fournier, L. Wang, O. V. Yazyev, and Q. Wu, Artifi- cial neural network approach to the analytic continuation problem, Phys. Rev. Lett.124, 056401 (2020)

    2020

  20. [20]

    Horak, J

    J. Horak, J. M. Pawlowski, J. Rodr´ ıguez-Quintero, J. Turnwald, J. M. Urban, N. Wink, and S. Zafeiropou- los, Reconstructing QCD spectral functions with Gaus- sian processes, Phys. Rev. D105, 036014 (2022), arXiv:2107.13464 [hep-ph]

    arXiv 2022

  21. [21]

    Bruno, L

    M. Bruno, L. Giusti, and M. Saccardi, Spectral densities from Euclidean lattice correlators via the Mellin trans- form, Phys. Rev. D111, 094515 (2025), arXiv:2407.04141 [hep-lat]

    arXiv 2025

  22. [22]

    Lawrence, Model-free spectral reconstruction via La- grange duality, arXiv:2408.11766 [hep-lat] (2024)

    S. Lawrence, Model-free spectral reconstruction via La- grange duality, arXiv:2408.11766 [hep-lat] (2024)

    arXiv 2024

  23. [23]

    Rattazzi, V

    R. Rattazzi, V. S. Rychkov, E. Tonni, and A. Vichi, Bounding scalar operator dimensions in 4D CFT, JHEP 12, 031, arXiv:0807.0004 [hep-th]

    Pith/arXiv arXiv

  24. [24]

    Poland, S

    D. Poland, S. Rychkov, and A. Vichi, The Confor- mal Bootstrap: Theory, Numerical Techniques, and Applications, Rev. Mod. Phys.91, 015002 (2019), arXiv:1805.04405 [hep-th]

    Pith/arXiv arXiv 2019

  25. [25]

    M. F. Paulos, J. Penedones, J. Toledo, B. C. van Rees, and P. Vieira, The S-matrix bootstrap. Part I: QFT in AdS, JHEP11, 133, arXiv:1607.06109 [hep-th]

    Pith/arXiv arXiv

  26. [26]

    Kruczenski, J

    M. Kruczenski, J. Penedones, and B. C. van Rees, Snowmass White Paper: S-matrix Bootstrap, arXiv:2203.02421 [hep-th] (2022)

    arXiv 2022

  27. [27]

    Karateev, S

    D. Karateev, S. Kuhn, and J. a. Penedones, Bootstrap- ping Massive Quantum Field Theories, JHEP07, 035, arXiv:1912.08940 [hep-th]

    arXiv 1912

  28. [28]

    H. Chen, A. L. Fitzpatrick, and D. Karateev, Bootstrap- ping 2dϕ 4 theory with Hamiltonian truncation data, JHEP02, 146, arXiv:2107.10286 [hep-th]

    arXiv

  29. [29]

    Osterwalder and R

    K. Osterwalder and R. Schrader, Axioms for Euclidean Green’s Functions. 2., Commun. Math. Phys.42, 281 (1975)

    1975

  30. [30]

    Montvay and G

    I. Montvay and G. M¨ unster,Quantum Fields on a Lat- tice, Cambridge Monographs on Mathematical Physics (Cambridge University Press, 1994)

    1994

  31. [31]

    Delcamp and A

    C. Delcamp and A. Tilloy, Computing the renormal- ization group flow of two-dimensionalϕ 4 theory with tensor networks, Phys. Rev. Res.2, 033278 (2020), arXiv:2003.12993 [cond-mat.str-el]

    arXiv 2020

  32. [32]

    Loinaz and R

    W. Loinaz and R. S. Willey, Monte Carlo simulation calculation of critical coupling constant for continuum phi**4 in two-dimensions, Phys. Rev. D58, 076003 (1998), arXiv:hep-lat/9712008

    Pith/arXiv arXiv 1998

  33. [33]

    Schaich and W

    D. Schaich and W. Loinaz, Improved lattice measurement of the critical coupling inϕ 4 2 theory, Phys. Rev. D79, 056008 (2009), arXiv:0902.0045 [hep-lat]

    Pith/arXiv arXiv 2009

  34. [34]

    Bosetti, B

    P. Bosetti, B. De Palma, and M. Guagnelli, Monte Carlo determination of the critical coupling inϕ 4 2 theory, Phys. Rev. D92, 034509 (2015), arXiv:1506.08587 [hep-lat]

    Pith/arXiv arXiv 2015

  35. [35]

    Bronzin, B

    S. Bronzin, B. De Palma, and M. Guagnelli, New Monte Carlo determination of the critical coupling inϕ 4 2 theory, Phys. Rev. D99, 034508 (2019), arXiv:1807.03381 [hep- lat]

    Pith/arXiv arXiv 2019

  36. [36]

    Lubin, O

    M. Lubin, O. Dowson, J. Dias Garcia, J. Huchette, B. Legat, and J. P. Vielma, JuMP 1.0: Recent im- provements to a modeling language for mathematical optimization, Mathematical Programming Computation 10.1007/s12532-023-00239-3 (2023)

    work page doi:10.1007/s12532-023-00239-3 2023

  37. [37]

    ApS,MOSEK Optimizer API for Julia 11.2.1(2025)

    M. ApS,MOSEK Optimizer API for Julia 11.2.1(2025)

    2025

  38. [38]

    C. Coey, L. Kapelevich, and J. P. Vielma, Solv- ing natural conic formulations with hypatia.jl (2021), arXiv:2005.01136 [math.OC]

    arXiv 2021

  39. [39]

    Alexandrouet al.(Extended Twisted Mass Collabo- ration (ETMC)), Probing the Energy-Smeared R Ratio Using Lattice QCD, Phys

    C. Alexandrouet al.(Extended Twisted Mass Collabo- ration (ETMC)), Probing the Energy-Smeared R Ratio Using Lattice QCD, Phys. Rev. Lett.130, 241901 (2023), arXiv:2212.08467 [hep-lat]

    arXiv 2023

  40. [40]

    Bergeron and A.-M

    D. Bergeron and A.-M. S. Tremblay, Algorithms for opti- mized maximum entropy and diagnostic tools for analytic continuation, Phys. Rev. E94, 023303 (2016)

    2016

  41. [41]

    R. Levy, J. LeBlanc, and E. Gull, Implementation of the maximum entropy method for analytic continua- tion, Computer Physics Communications215, 149–155 (2017)

    2017

  42. [42]

    G. J. Kraberger, R. Triebl, M. Zingl, and M. Aichhorn, Maximum entropy formalism for the analytic continua- tion of matrix-valued green’s functions, Phys. Rev. B96, 155128 (2017)

    2017

  43. [43]

    Otsuki, M

    J. Otsuki, M. Ohzeki, H. Shinaoka, and K. Yoshimi, Sparse modeling approach to analytical continuation of imaginary-time quantum monte carlo data, Phys. Rev. E 95, 061302(R) (2017)

    2017

  44. [44]

    Fei, C.-N

    J. Fei, C.-N. Yeh, and E. Gull, Nevanlinna analytical con- tinuation, Phys. Rev. Lett.126, 056402 (2021)

    2021

  45. [45]

    Kaufmann and K

    J. Kaufmann and K. Held, ana cont: Python package for analytic continuation, Computer Physics Communi- cations282, 108519 (2023)

    2023

  46. [46]

    Mutzel and A

    S. Mutzel and A. Tilloy, Extracting quantum field theory dynamics from an approximate ground state, arXiv:2512.19594 [quant-ph] (2025)

    arXiv 2025

  47. [47]

    Abbott, S

    R. Abbott, S. Fields, W. I. Jay, P. Oare, and M. Sac- cardi, The causal bootstrap: Bounding smeared spectral functions from non-perturbative euclidean data (2026), arXiv:2605.20509 [hep-lat]

    Pith/arXiv arXiv 2026

  48. [48]

    Boyd and L

    S. Boyd and L. Vandenberghe,Convex Optimization (Cambridge University Press, 2004)

    2004

  49. [49]

    Nocedal and S

    J. Nocedal and S. J. Wright,Numerical optimization, 2nd ed., Springer series in operations research and financial engineering (Springer, 2006)

    2006

  50. [50]

    V. A. Marˇ cenko and L. A. Pastur, Distribution of eigen- values for some sets of random matrices, Mathematics of the USSR-Sbornik1, 457 (1967)

    1967

  51. [51]

    Yoon, Y.-C

    B. Yoon, Y.-C. Jang, C. Jung, and W. Lee, Covariance fitting of highly correlated data in lattice QCD, J. Korean Phys. Soc.63, 145 (2013), arXiv:1101.2248 [hep-lat]

    Pith/arXiv arXiv 2013

  52. [52]

    Bruno and R

    M. Bruno and R. Sommer, On fits to correlated and auto- correlated data, Comput. Phys. Commun.285, 108643 (2023), arXiv:2209.14188 [hep-lat]. 11 APPENDIX Appendix A: Derivation of dual problem from Lagrange duality In this appendix we derive the dual optimization prob- lem (21) from standard Lagrange duality. In the follow- ing we use the notation (f 1, f...

    arXiv 2023