Pith. sign in

REVIEW 2 major objections 131 references

Optimizing the two free parameters of 3D Kronecker point sets beats prior L∞ star-discrepancy records for every size of at least 500 points, and irace finds parameters that set new records across whole ranges of sizes.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

T0 review · grok-4.5

2026-07-13 14:50 UTC pith:AWUR7AN5

load-bearing objection We only have the abstract for the Kronecker/irace paper; the supplied full text is an unrelated Λ(1405) nuclear-physics preprint, so the SOTA claims cannot be audited. the 2 major comments →

arxiv 2604.00786 v2 pith:AWUR7AN5 submitted 2026-04-01 cs.NE

Finding Low Star Discrepancy 3D Kronecker Point Sets Using Algorithm Configuration Techniques

classification cs.NE MSC 11K3865C0568T20
keywords star discrepancyKronecker sequencesalgorithm configurationiracelow-discrepancy point setsquasi-Monte Carloexperimental designBayesian optimization
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

Low L∞ star discrepancy means a point set fills the unit cube as evenly as possible. Such sets are used as experimental designs, as starting designs for Bayesian optimization, and for quasi-Monte Carlo integration. Classical infinite sequences (Sobol', Halton, Hammersley) are known to be far from optimal once the number of points is fixed. This paper asks how much one classical parametric family—the three-dimensional Kronecker construction—can be improved by size-specific tuning of its two free parameters. Direct optimization already produces sets better than the previous state of the art for every n ≥ 500. Feeding the same construction into the automated configurator irace then yields parameter values that establish new record discrepancy figures for contiguous ranges of set sizes. The practical payoff is better fixed-n designs for the applications above without inventing an entirely new construction.

Core claim

Size-specific optimization of the two configurable parameters of the three-dimensional Kronecker construction produces point sets whose L∞ star discrepancy outperforms the previous state of the art for every set size of at least 500 points; parameters found by irace further deliver new state-of-the-art values across whole ranges of n.

What carries the argument

The two-parameter 3D Kronecker construction (a simple arithmetic lattice determined by two real generators) searched by the irace algorithm-configuration procedure to minimize L∞ star discrepancy at target cardinalities.

Load-bearing premise

That the two-parameter Kronecker family, once tuned, is not already dominated by other size-specific constructions in the literature and that the reported comparisons use the same discrepancy algorithm, precision and baseline sets as the prior state of the art.

What would settle it

Recompute the L∞ star discrepancy of the published Kronecker parameter sets with an independent high-precision implementation and compare them, for each n ≥ 500, against the previously best-known point sets under identical evaluation; any prior set that is better falsifies the claim.

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

If this is right

  • For any fixed n ≥ 500 in three dimensions, the tuned Kronecker sets can replace classical Sobol'/Halton/Hammersley designs whenever lower star discrepancy is the goal.
  • Algorithm configuration becomes a practical, off-the-shelf route to improve existing parametric constructions when the application cares about fixed finite n rather than asymptotic rates.
  • New numerical tables of best-known 3D star-discrepancy values become available for contiguous ranges of set sizes.
  • The same size-specific tuning strategy can be applied to other low-dimensional parametric families used in experimental design and quasi-Monte Carlo.

Where Pith is reading between the lines

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

  • The same irace pipeline could be run on higher-dimensional Kronecker or lattice rules whose few continuous parameters still control uniformity.
  • If the advantage survives under a common discrepancy code base, published “best-known” tables for 3D star discrepancy need systematic revision for n ≥ 500.
  • Fixed-n optimization may matter more than convergence-rate design for the majority of practical budgets that never exceed a few thousand points.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 0 minor

Summary. The submitted abstract claims that size-specific optimization of the two free parameters of the 3-dimensional Kronecker construction, performed with the algorithm configurator irace, produces point sets whose L∞ star discrepancy improves on published state-of-the-art values for all n ≥ 500 and for contiguous ranges of set sizes. The body of the manuscript supplied for review, however, is an entirely different work (off-shell unitarized chiral EFT for the S = −1 meson–baryon system and K−p / πΣ femtoscopic correlation functions). Consequently none of the claimed constructions, discrepancy tables, baseline comparisons, irace budgets, or statistical protocols appear in the document under review.

Significance. If the abstract’s claims were substantiated by a matching manuscript, the result would be of clear practical interest to quasi-Monte Carlo, design of experiments and Bayesian optimization, because it would demonstrate that a classical two-parameter family can be made competitive with modern size-specific constructions simply by automated configuration. That potential significance cannot be assessed from the material actually provided.

major comments (2)
  1. The full manuscript text is unrelated to the title and abstract (arXiv:2604.00786). It is instead the nuclear-physics preprint on off-shell chiral dynamics of the Λ(1405) and K−p femtoscopy (arXiv:2604.00791). No Kronecker generators, no L∞ star-discrepancy values, no irace runs, no baseline tables and no evaluation protocol are present. The central claim is therefore completely unauditable.
  2. Because the supplied body contains none of the numerical or methodological content required to verify the SOTA comparisons asserted in the abstract, it is impossible to check fairness of baselines, identity of the discrepancy algorithm, precision, or statistical significance of the reported improvements for n ≥ 500.

Circularity Check

0 steps flagged

No circularity found: abstract describes empirical irace configuration of Kronecker parameters against an external L∞ star-discrepancy objective; provided full text is an unrelated nuclear-physics manuscript and cannot support any reduction.

full rationale

The target paper (arXiv:2604.00786) is available only via its abstract. That abstract frames a standard algorithm-configuration study: two free parameters of the 3D Kronecker construction are optimized with irace so that the generated point sets minimize the externally defined L∞ star discrepancy, and the resulting values are compared to prior SOTA. Nothing in the abstract equates the reported discrepancy to the fitted generators by definition, renames a known result, or imports a uniqueness theorem from the same authors. The residual methodological risk that parameters were tuned on the same set sizes later used for SOTA claims is ordinary for configuration work and is not circularity under the stated criteria. The CACHEABLE full-manuscript block is an entirely different preprint (off-shell chiral dynamics of Λ(1405) / K−p femtoscopy, arXiv:2604.00791) and contains none of the Kronecker construction, irace runs, discrepancy tables, or baselines; it therefore supplies no equations or self-citations that could exhibit a circular reduction for the claimed result. With no quotable reduction available, the honest finding is score 0 and an empty steps list.

Axiom & Free-Parameter Ledger

2 free parameters · 4 axioms · 0 invented entities

Abstract-only review of a configuration paper. Load-bearing ingredients are the definition of L∞ star discrepancy, the two-parameter Kronecker construction, the claim that classical sequences are suboptimal at fixed n, and the use of irace as a black-box configurator. No new physical entities. Free parameters are precisely the two Kronecker generators (and any irace meta-parameters left unspecified).

free parameters (2)
  • Kronecker generators (two real parameters in 3D)
    The abstract states the construction has exactly two configurable parameters that are optimized; their fitted values are the object of the irace search and are not given in the abstract.
  • irace configuration budget and training instance set (sizes n)
    Any automated configurator depends on which n are used for training, the evaluation budget, and irace's own hyperparameters; none are specified in the abstract but the SOTA claims rest on them.
axioms (4)
  • domain assumption L∞ star discrepancy is the right uniformity measure for the target applications (DoE, BO initial designs, QMC).
    Stated in the opening of the abstract as the evaluation criterion; alternative measures (L2, wrap-around, etc.) are not discussed.
  • domain assumption Classical constructions (Sobol', Halton, Hammersley) can be substantially outperformed at fixed n, so size-specific optimization is worthwhile.
    Cited as 'recent work' in the abstract; the paper's motivation rests on that premise.
  • domain assumption The Kronecker construction in 3D is fully described by two real parameters whose optimization is sufficient to reach new SOTA.
    Abstract treats the two-parameter family as the search space; no other degrees of freedom are mentioned.
  • ad hoc to paper irace is an adequate global search method for this continuous two-parameter space.
    Choice of configurator is methodological; alternatives (CMA-ES, grid search, Bayesian optimization of the generators) are not compared in the abstract.

pith-pipeline@v1.1.0-grok45 · 16749 in / 2847 out tokens · 32657 ms · 2026-07-13T14:50:01.595450+00:00 · methodology

0 comments
read the original abstract

The L infinity star discrepancy is a measure for how uniformly a point set is distributed in a given space. Point sets of low star discrepancy are used as designs of experiments, as initial designs for Bayesian optimization algorithms, for quasi-Monte Carlo integration methods, and many other applications. Recent work has shown that classical constructions such as Sobol', Halton, or Hammersley sequences can be outperformed by large margins when considering point sets of fixed sizes rather than their convergence behavior. These results, highly relevant to the aforementioned applications, raise the question of how much existing constructions can be improved through size-specific optimization. In this work, we study this question for the so-called Kronecker construction. Focusing on the 3-dimensional setting, we show that optimizing the two configurable parameters of its construction yields point sets outperforming the state-of-the-art value for sets of at least 500 points. Using the algorithm configuration technique irace, we then derive parameters that yield new state-of-the-art discrepancy values for whole ranges of set sizes.

discussion (0)

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

Reference graph

Works this paper leans on

131 extracted references · 85 linked inside Pith

  1. [1]

    D. J. Gross and F. Wilczek, Phys. Rev. Lett.30, 1343 (1973)

  2. [2]

    H. D. Politzer, Phys. Rev. Lett.30, 1346 (1973)

  3. [3]

    D. J. Gross and F. Wilczek, Phys. Rev. D8, 3633 (1973)

  4. [4]

    Grosset al., Eur

    F. Grosset al., Eur. Phys. J. C83, 1125 (2023), arXiv:2212.11107 [hep-ph]

  5. [5]

    C. J. Batty, E. Friedman, and A. Gal, Phys. Rept.287, 385 (1997)

  6. [6]

    Hyodo and D

    T. Hyodo and D. Jido, Prog. Part. Nucl. Phys.67, 55 (2012), arXiv:1104.4474 [nucl-th]

  7. [7]

    A. Gal, E. V . Hungerford, and D. J. Millener, Rev. Mod. Phys. 88, 035004 (2016), arXiv:1605.00557 [nucl-th]

  8. [8]

    Curceanuet al., Rev

    C. Curceanuet al., Rev. Mod. Phys.91, 025006 (2019)

  9. [9]

    Hyodo and M

    T. Hyodo and M. Niiyama, Prog. Part. Nucl. Phys.120, 103868 (2021), arXiv:2010.07592 [hep-ph]

  10. [10]

    R. H. Dalitz and S. F. Tuan, Phys. Rev. Lett.2, 425 (1959)

  11. [11]

    J. K. Kim, Phys. Rev. Lett.14, 29 (1965)

  12. [12]

    C. G. Callan, Jr. and I. R. Klebanov, Nucl. Phys. B262, 365 (1985)

  13. [13]

    R. H. Dalitz, T. C. Wong, and G. Rajasekaran, Phys. Rev.153, 1617 (1967)

  14. [14]

    E. A. Veit, B. K. Jennings, R. C. Barrett, and A. W. Thomas, Phys. Lett. B137, 415 (1984)

  15. [15]

    E. A. Veit, B. K. Jennings, A. W. Thomas, and R. C. Barrett, Phys. Rev. D31, 1033 (1985)

  16. [16]

    B. K. Jennings, Phys. Lett. B176, 229 (1986)

  17. [17]

    P. J. Fink, Jr., G. He, R. H. Landau, and J. W. Schnick, Phys. Rev. C41, 2720 (1990)

  18. [18]

    Capstick and N

    S. Capstick and N. Isgur, Phys. Rev. D34, 2809 (1986)

  19. [19]

    Klempt and J.-M

    E. Klempt and J.-M. Richard, Rev. Mod. Phys.82, 1095 (2010), arXiv:0901.2055 [hep-ph]

  20. [20]

    Weinberg, Phys

    S. Weinberg, Phys. Lett. B251, 288 (1990)

  21. [21]

    Kaiser, P

    N. Kaiser, P. B. Siegel, and W. Weise, Nucl. Phys. A594, 325 (1995), arXiv:nucl-th/9505043

  22. [22]

    Kaiser, T

    N. Kaiser, T. Waas, and W. Weise, Nucl. Phys. A612, 297 (1997), arXiv:hep-ph/9607459

  23. [23]

    Oset and A

    E. Oset and A. Ramos, Nucl. Phys. A635, 99 (1998), arXiv:nucl-th/9711022

  24. [24]

    J. A. Oller, E. Oset, and A. Ramos, Prog. Part. Nucl. Phys.45, 157 (2000), arXiv:hep-ph/0002193

  25. [25]

    J. A. Oller, Prog. Part. Nucl. Phys.110, 103728 (2020), arXiv:1909.00370 [hep-ph]

  26. [26]

    Mai, Eur

    M. Mai, Eur. Phys. J. ST230, 1593 (2021), arXiv:2010.00056 [nucl-th]

  27. [27]

    J. A. Oller and U.-G. Meißner, Phys. Lett. B500, 263 (2001), 10 arXiv:hep-ph/0011146

  28. [28]

    D. Jido, J. A. Oller, E. Oset, A. Ramos, and U.-G. Meißner, Nucl. Phys. A725, 181 (2003), arXiv:nucl-th/0303062

  29. [29]

    Meißner, Symmetry12, 981 (2020), arXiv:2005.06909 [hep-ph]

    U.-G. Meißner, Symmetry12, 981 (2020), arXiv:2005.06909 [hep-ph]

  30. [30]

    Mai, U.-G

    M. Mai, U.-G. Meißner, and C. Urbach, Phys. Rept.1001, 1 (2023), arXiv:2206.01477 [hep-ph]

  31. [31]

    Geng, M.-Z

    L.-S. Geng, M.-Z. liu, and J.-M. Xie, in15th International Conference on Hypernuclear and Strange Particle Physics (2026) arXiv:2602.17190 [hep-ph]

  32. [32]

    He, X.-H

    Y .-B. He, X.-H. Liu, L.-S. Geng, F.-K. Guo, and J.-J. Xie, Phys. Rev. D113, L051501 (2026)

  33. [33]

    Bulavaet al.(Baryon Scattering (BaSc)), Phys

    J. Bulavaet al.(Baryon Scattering (BaSc)), Phys. Rev. Lett. 132, 051901 (2024), arXiv:2307.10413 [hep-lat]

  34. [34]

    Bulavaet al.(Baryon Scattering (BaSc)), Phys

    J. Bulavaet al.(Baryon Scattering (BaSc)), Phys. Rev. D109, 014511 (2024), arXiv:2307.13471 [hep-lat]

  35. [35]

    Borasoy, R

    B. Borasoy, R. Nissler, and W. Weise, Eur. Phys. J. A25, 79 (2005), arXiv:hep-ph/0505239

  36. [36]

    Borasoy, U.-G

    B. Borasoy, U.-G. Meißner, and R. Nissler, Phys. Rev. C74, 055201 (2006), arXiv:hep-ph/0606108

  37. [37]

    J. A. Oller, J. Prades, and M. Verbeni, Phys. Rev. Lett.95, 172502 (2005), arXiv:hep-ph/0508081

  38. [38]

    J. A. Oller, M. Verbeni, and J. Prades, JHEP09, 079 (2006), arXiv:hep-ph/0608204

  39. [39]

    Ikeda, T

    Y . Ikeda, T. Hyodo, and W. Weise, Nucl. Phys. A881, 98 (2012), arXiv:1201.6549 [nucl-th]

  40. [40]

    Guo and J

    Z.-H. Guo and J. A. Oller, Phys. Rev. C87, 035202 (2013), arXiv:1210.3485 [hep-ph]

  41. [41]

    Mai and U.-G

    M. Mai and U.-G. Meißner, Nucl. Phys. A900, 51 (2013), arXiv:1202.2030 [nucl-th]

  42. [42]

    Ramos, A

    A. Ramos, A. Feijoo, and V . K. Magas, Nucl. Phys. A954, 58 (2016), arXiv:1605.03767 [nucl-th]

  43. [43]

    Sadasivan, M

    D. Sadasivan, M. Mai, M. D¨oring, U.-G. Meißner, F. Amorim, J. P. Klucik, J.-X. Lu, and L.-S. Geng, Front. Phys.11, 1139236 (2023), arXiv:2212.10415 [nucl-th]

  44. [44]

    Lu, L.-S

    J.-X. Lu, L.-S. Geng, M. Doering, and M. Mai, Phys. Rev. Lett.130, 071902 (2023), arXiv:2209.02471 [hep-ph]

  45. [45]

    R ´evai, Few Body Syst.59, 49 (2018), arXiv:1711.04098 [nucl-th]

    J. R ´evai, Few Body Syst.59, 49 (2018), arXiv:1711.04098 [nucl-th]

  46. [46]

    Nieves and E

    J. Nieves and E. Ruiz Arriola, Phys. Rev. D64, 116008 (2001), arXiv:hep-ph/0104307

  47. [47]

    Garcia-Recio, J

    C. Garcia-Recio, J. Nieves, E. Ruiz Arriola, and M. J. Vi- cente Vacas, Phys. Rev. D67, 076009 (2003), arXiv:hep- ph/0210311

  48. [48]

    Morimatsu and K

    O. Morimatsu and K. Yamada, Phys. Rev. C100, 025201 (2019), arXiv:1903.12380 [hep-ph]

  49. [49]

    J. A. Oller and E. Oset, Nucl. Phys. A620, 438 (1997), [Erratum: Nucl.Phys.A 652, 407–409 (1999)], arXiv:hep- ph/9702314

  50. [50]

    Sadasivan, M

    D. Sadasivan, M. Mai, and M. D ¨oring, Phys. Lett. B789, 329 (2019), arXiv:1805.04534 [nucl-th]

  51. [51]

    Xie, J.-X

    J.-M. Xie, J.-X. Lu, L.-S. Geng, and B.-S. Zou, Phys. Rev. D 108, L111502 (2023), arXiv:2307.11631 [hep-ph]

  52. [52]

    Xie, J.-X

    J.-M. Xie, J.-X. Lu, L.-S. Geng, and B.-S. Zou, EPJ Web Conf.303, 01011 (2024), arXiv:2312.17287 [hep-ph]

  53. [53]

    Xie, J.-X

    J.-M. Xie, J.-X. Lu, L.-S. Geng, and B.-S. Zou, PoSCD2024, 005 (2026), arXiv:2504.03392 [hep-ph]

  54. [54]

    F.-K. Guo, Y . Kamiya, M. Mai, and U.-G. Meißner, Phys. Lett. B846, 138264 (2023), arXiv:2308.07658 [hep-ph]

  55. [55]

    Zhuang, R

    Z. Zhuang, R. Molina, J.-X. Lu, and L.-S. Geng, Sci. Bull.70, 1953 (2025), arXiv:2405.07686 [hep-ph]

  56. [56]

    Pittler, M

    F. Pittler, M. Mai, U.-G. Meißner, R. F. Ferguson, P. Hurck, D. G. Ireland, and B. McKinnon, Phys. Rev. D112, 074037 (2025), arXiv:2507.14283 [hep-ph]

  57. [57]

    Nieves and E

    J. Nieves and E. Ruiz Arriola, Phys. Lett. B455, 30 (1999), arXiv:nucl-th/9807035

  58. [58]

    Nieves and E

    J. Nieves and E. Ruiz Arriola, Nucl. Phys. A679, 57 (2000), arXiv:hep-ph/9907469

  59. [59]

    Altenbuchinger and L.-S

    M. Altenbuchinger and L.-S. Geng, Phys. Rev. D89, 054008 (2014), arXiv:1310.5224 [hep-ph]

  60. [60]

    Nieves and E

    J. Nieves and E. Ruiz Arriola, Phys. Rev. D63, 076001 (2001), arXiv:hep-ph/0008034

  61. [61]

    P. C. Bruns, M. Mai, and U.-G. Meißner, Phys. Lett. B697, 254 (2011), arXiv:1012.2233 [nucl-th]

  62. [62]

    A. Dote, T. Hyodo, and W. Weise, Nucl. Phys. A804, 197 (2008), arXiv:0802.0238 [nucl-th]

  63. [63]

    A. Dote, T. Hyodo, and W. Weise, Phys. Rev. C79, 014003 (2009), arXiv:0806.4917 [nucl-th]

  64. [64]

    Ikeda, H

    Y . Ikeda, H. Kamano, and T. Sato, Prog. Theor. Phys.124, 533 (2010), arXiv:1004.4877 [nucl-th]

  65. [65]

    Bayar, J

    M. Bayar, J. Yamagata-Sekihara, and E. Oset, Phys. Rev. C 84, 015209 (2011), arXiv:1102.2854 [hep-ph]

  66. [66]

    Acharyaet al.(ALICE), Phys

    S. Acharyaet al.(ALICE), Phys. Rev. Lett.124, 092301 (2020), arXiv:1905.13470 [nucl-ex]

  67. [67]

    Acharyaet al.(ALICE), Eur

    S. Acharyaet al.(ALICE), Eur. Phys. J. C83, 340 (2023), arXiv:2205.15176 [nucl-ex]

  68. [68]

    Kamiya, T

    Y . Kamiya, T. Hyodo, K. Morita, A. Ohnishi, and W. Weise, Phys. Rev. Lett.124, 132501 (2020), arXiv:1911.01041 [nucl- th]

  69. [69]

    Encarnaci ´on, A

    P. Encarnaci ´on, A. Feijoo, V . M. Sarti, and A. Ramos, Phys. Rev. D111, 114013 (2025), arXiv:2412.20880 [hep-ph]

  70. [70]

    Morita, A

    K. Morita, A. Ohnishi, F. Etminan, and T. Hatsuda, Phys. Rev. C94, 031901 (2016), [Erratum: Phys.Rev.C 100, 069902 (2019)], arXiv:1605.06765 [hep-ph]

  71. [71]

    Morita, S

    K. Morita, S. Gongyo, T. Hatsuda, T. Hyodo, Y . Kamiya, and A. Ohnishi, Phys. Rev. C101, 015201 (2020), arXiv:1908.05414 [nucl-th]

  72. [72]

    Liu, K.-W

    Z.-W. Liu, K.-W. Li, and L.-S. Geng, Chin. Phys. C47, 024108 (2023), arXiv:2201.04997 [hep-ph]

  73. [73]

    Vidana, A

    I. Vidana, A. Feijoo, M. Albaladejo, J. Nieves, and E. Oset, Phys. Lett. B846, 138201 (2023), arXiv:2303.06079 [hep-ph]

  74. [74]

    Ikeno, G

    N. Ikeno, G. Toledo, and E. Oset, Phys. Lett. B847, 138281 (2023), arXiv:2305.16431 [hep-ph]

  75. [75]

    J. M. Torres-Rincon, `A. Ramos, and L. Tolos, Phys. Rev. D 108, 096008 (2023), arXiv:2307.02102 [hep-ph]

  76. [76]

    Liu, J.-X

    Z.-W. Liu, J.-X. Lu, and L.-S. Geng, Phys. Rev. D107, 074019 (2023), arXiv:2302.01046 [hep-ph]

  77. [77]

    Liu, J.-X

    Z.-W. Liu, J.-X. Lu, M.-Z. Liu, and L.-S. Geng, Phys. Rev. D 108, L031503 (2023), arXiv:2305.19048 [hep-ph]

  78. [78]

    Molina, Z.-W

    R. Molina, Z.-W. Liu, L.-S. Geng, and E. Oset, Eur. Phys. J. C84, 328 (2024), arXiv:2312.11993 [hep-ph]

  79. [79]

    Albaladejo, A

    M. Albaladejo, A. Feijoo, J. Nieves, E. Oset, and I. Vida ˜na, Phys. Rev. D110, 114052 (2024), arXiv:2410.08880 [hep-ph]

  80. [80]

    V . M. Sarti, A. Feijoo, I. Vida˜na, A. Ramos, F. Giacosa, T. Hy- odo, and Y . Kamiya, Phys. Rev. D110, L011505 (2024), arXiv:2309.08756 [hep-ph]

Showing first 80 references.