Pith. sign in

REVIEW 2 major objections 6 minor 1 cited by

For very massive LISA binaries, neglecting a higher harmonic can land the inferred source on the wrong side of the sky.

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 →

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

T0 review · deepseek-v4-flash

2026-08-03 03:02 UTC pith:BCCQZONE

load-bearing objection Solid, honest extension of Paper 1 showing neglected higher harmonics can badly bias LISA MBHB parameter estimation, including a striking sky-mislocalization example; the quantitative maps are model-dependent, but the qualitative message holds. the 2 major comments →

arxiv 2602.09088 v2 pith:BCCQZONE submitted 2026-02-09 gr-qc astro-ph.COastro-ph.HE

Systematic biases in parameter estimation on LISA binaries. II. The effect of excluding higher harmonics for spin-aligned, high-mass binaries

Sophia Yi , Francesco Iacovelli , Emanuele Berti , Rohit S. Chandramouli , Sylvain Marsat , Digvijay Wadekar , Nicol\'as Yunes This is my paper
classification gr-qc astro-ph.COastro-ph.HE PACS 04.30.-w04.80.Nn
keywords LISAmassive black hole binarieshigher-order modessystematic biasesparameter estimationsky localizationspin-aligned binarieslikelihood 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.

This paper claims that LISA's loudest sources—massive black hole binaries with total mass above about five million solar masses—can have higher-order waveform modes that out-rank the usually dominant (2,2) mode, and that dropping just one such mode from the analysis can push parameter estimation onto the wrong sky position. The bias magnitude depends strongly on the progenitor spins, and for the shortest, heaviest signals even small waveform errors can make the sky localization confidently wrong. If correct, LISA data analysis for heavy binaries must include higher modes and treat sky location as possibly multimodal. The paper also shows that a fast likelihood-optimization scheme can predict these biases and per-sky-octant maxima far more cheaply than full Bayesian runs.

Core claim

For MBHBs with detector-frame total mass greater than about 5×10^6 M_sun and mass ratio greater than 5, the (2,2) quadrupole is no longer guaranteed to dominate; (3,3) and (4,4) harmonics can contribute comparable or larger SNR, especially away from face-on inclination and depending on the aligned-spin configuration. Injecting full waveforms and recovering with the (3,2) mode omitted, the authors show systematic biases can exceed twice the statistical error for a significant fraction of detectable events at redshift below about 2.5, and for an M=10^7 M_sun, q=1.1, ι=π/3 example the posterior maximum moves to the 'reflected' sky octant—an entirely wrong position—even though the local maxima i

What carries the argument

The analysis decomposes the gravitational-wave signal into spin-weighted spherical harmonics (ℓ,m), keeping the (2,2), (2,1), (3,3), (3,2), (4,4) modes, and computes the SNR inner product including cross-terms between harmonics; mode ranking is driven by total mass, mass ratio, inclination, and spins, with galactic-binary confusion noise altering the ranking near M~10^7 M_sun. To predict biases it uses direct likelihood maximization upgraded with a global optimization step, a reparametrization to less-correlated variables (log mass ratio, an effective spin combination, antisymmetric spin, cosine inclination, chirp distance), bounds derived from the local Gaussian covariance, and octant-restr

Load-bearing premise

The quantitative claims rest on the injected waveform model (IMRPhenomXHM, original release) faithfully representing the true higher-mode amplitudes and phases; the model's own artifacts force the authors to exclude some spin regions, so if this model is wrong in the strengths of (3,3) or (3,2), the predicted bias maps and octant flips may not occur for real signals.

What would settle it

For the M=10^7 M_sun, q=1.1, ι=π/3, zero-spin event, inject a waveform from an independent numerical-relativity-calibrated model and recover with a template omitting only the (3,2) mode; if the maximum likelihood stays in the true sky octant rather than the reflected one, the octant-switching claim fails. A second check: map the boundary where (3,3) and (4,4) out-rank (2,2) in SNR using an independent waveform family; if the boundary disappears, the hierarchy-reversal result is model-dependent.

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

If this is right

  • For MBHBs with total mass above about 5×10^6 M_sun and mass ratio above 5, templates that omit modes beyond (2,2) will produce parameter biases that exceed statistical errors, so these modes must be included in LISA parameter estimation.
  • Sky localization for the heaviest events must be treated as potentially multimodal: the maximum-likelihood octant can differ from the true one even when the posterior within an octant is narrow.
  • Under benchmark population models, roughly 6–22% of detectable MBHBs lie at redshift below 2.5, and the bias maps imply a comparable fraction could suffer significant bias from a single neglected mode.
  • The improved likelihood optimization reproduces the biases and per-octant maxima obtained from full Bayesian sampling using a small fraction of the evaluations, making it practical to map biases across masses, mass ratios, inclinations, and spins.
  • For total mass above about 10^8 M_sun, differences between candidate sky octants become indistinguishable from noise fluctuations, so LISA alone may not localize the source to a specific octant.

Where Pith is reading between the lines

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

  • The same octant-switching mechanism should apply to other systematic waveform errors, not just missing modes; waveform-accuracy requirements for the loudest events may need to be set by sky-localization stability rather than by template-match thresholds.
  • The finding that a sub-noise-threshold log-likelihood difference can correspond to a completely different sky position suggests that the usual criterion for 'acceptable' systematic bias may need to be replaced by a multimodality-aware criterion for LISA science.
  • If the mode-hierarchy reversal persists in precessing or eccentric binaries (not covered here), lower-mode-only search banks could miss the true posterior peak entirely, which would affect global-fit pipelines that subtract individual sources.
  • The zero-noise assumption is optimistic in one direction and pessimistic in another: real noise could make the wrong octant look even better, so multi-octant follow-up may be needed for verification of heavy MBHB detections.

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 / 6 minor

Summary. This paper extends the authors' earlier work (Paper I) to higher total masses (up to 10^8 M_sun) and to aligned-spin, nonprecessing binaries. Using zero-noise injections with IMRPhenomXHM, the authors compare three ways of estimating systematic bias from neglected higher harmonics: full Bayesian PE with ptemcee, the Cutler-Vallisneri linear-signal approximation, and direct likelihood optimization (improved with dual annealing, reparametrization, and Fisher-informed priors). The main results are: (i) for M ≳ 5×10^6 M_sun and q ≳ 5, the (2,2) mode is not always the dominant harmonic, with mass ratio, inclination, and spin strongly affecting the mode hierarchy; (ii) the redshift below which neglecting a subdominant mode, e.g. (3,2), biases intrinsic parameters by more than 2σ varies strongly with spin; (iii) for a M = 10^7 M_sun, q = 1.1 event, omitting the (3,2) mode moves the maximum log-likelihood to a reflected sky octant; and (iv) for M = 10^8 M_sun, the octants become nearly indistinguishable. The paper is careful to gray out regions of low waveform-calibration confidence and to exclude a region affected by a known (3,3) artifact in IMRPhenomXHM v122019.

Significance. If the results hold, they have practical implications for LISA MBHB analyses: higher harmonics are not a small correction for the loudest high-mass events, and sky localization can be multimodal, with a bias-driven octant preference. The improved likelihood optimization pipeline is a useful, fast complement to full Bayesian PE, and the paper validates it against ptemcee on several events. There are no fitted constants in the central bias estimates. The honest handling of waveform-model limitations—graying out |χ|>0.9 regions, excluding the χ1 < -0.38 region in Fig. 7, and cross-checking the (2,1) amplitude dips against IMRPhenomHM and SEOBNRv5HM_ROM—is a clear strength. The main barrier to full acceptance is that the headline sky-mislocalization result currently rests on a single waveform-model version, without an independent cross-check.

major comments (2)
  1. [Sec. V, Figs. 8–9; Appendix B.2] The central claim of confident sky mislocalization is demonstrated for a single event (M=10^7 M_sun, q=1.1, nonspinning, i=π/3) using IMRPhenomXHM v122019 for both injection and recovery. Appendix B.2 documents a known unphysical (3,3) feature in this version, corrected in v122022. Although that artifact is not localized at the Fig. 8 event, the bias is controlled by the relative amplitudes of the (3,2), (3,3), and (4,4) harmonics, and no independent-waveform check is provided for this event. Please repeat the octant-restricted likelihood comparison of Fig. 9 with v122022 or SEOBNRv5HM_ROM and either report that the octant switch persists, or reframe the abstract and conclusions to make the result conditional on the specific approximant.
  2. [Sec. IV, Fig. 7; Appendix B] The spin-dependent critical-redshift map is a central result, but it uses the same v122019 model and explicitly excludes a region where the model is pathological (χ1 < -0.38 for q=1.1). The residual risk that similar model artifacts affect other (q, χ) combinations is not discussed. Please add a brief model-robustness statement: for a few representative points, compare v122019 against v122022 or SEOBNRv5HM_ROM and state whether the qualitative spin trends in Fig. 7 survive. This would separate physical spin effects from waveform-model artifacts.
minor comments (6)
  1. [Algorithm 1] The pseudo-code contains a typo: 'θ⋆e ← θ⋆e ← θ⋆(e)e,n (lnL(e)max)' should be a single assignment. Also 'len[all(Δθ⋆e,n ≤ 10%)]' is unclear; define the convergence criterion more formally.
  2. [Eq. (3)] The transformed variables are defined in terms of χ+ and χ-, but these combinations are defined only in the following sentence. Define χ± before Eq. (3), and state explicitly that geometric units G=c=1 are used so that the 'chirp distance' combination has the intended scaling.
  3. [Fig. 5] The caption says 'the systematic biases on two MBHB parameters', but the axes show 11 parameters. This appears to mean 'two MBHB events'.
  4. [Abstract / Conclusions] The phrase 'heaviest, and therefore shortest' is imprecise: the confident octant-switching demonstration is at 10^7 M_sun, while at 10^8 M_sun the octants are nearly indistinguishable (Table I). Rephrase to distinguish the two regimes.
  5. [Appendix B.2 / Sec. IV] The v122019-vs-v122022 distinction is discussed only in an appendix, but it underlies the main results. Consider summarizing the version dependence in Sec. IV as well, since it is directly relevant to the grayed-out and excluded regions.
  6. [Abstract / Sec. IIIB] The phrase 'predict these effects' is stronger than what the method establishes: the validation shows that the optimizer recovers the PE maximum on the same likelihood surface. Suggest 'estimate' or 'recover' rather than 'predict'.

Circularity Check

0 steps flagged

No significant circularity: the bias, mode-ordering, and octant-switching results are computed mismatch-simulation outputs, not self-referential derivations.

full rationale

The paper's central chain is a controlled zero-noise injection/recovery experiment: Eq. (1) defines the SNR, the mode hierarchies and cross terms are computed directly from IMRPhenomXHM, full signals are injected, reduced-mode templates are used in recovery, and the likelihood is maximized to map systematic biases. No fitted constant or parameter is calibrated against the claimed bias maps; the octant-switching result (Sec. V, Figs. 8-9) is a direct comparison of lnL_max across eight near-degenerate sky octants, so statements such as 'the lnL_max found in the (-1,0) octant is indeed slightly higher, with the difference ... Delta lnL_max = 5.859' are computed properties of the waveform pair, not assumed inputs. Validation against ptemcee on the same likelihood surface is a convergence check, not an independent physical prediction; that is a methodological limitation, not circularity. Self-citations (Paper 1; Marsat et al. 2021; Marsat in prep.) supply the software/response framework and context, but the new high-mass/spin results are produced and internally cross-checked here. The disclosed model-dependence caveats - Sec. IV graying out |chi|>0.9 ('less confident in the results shown'), Appendix B.2 excluding chi1<-0.38 because of the version-122019 (3,3)-mode artifact, and the stated assumption that IMRPhenomXHM does not 'substantially overestimate or underestimate the contribution of higher-order modes' - are genuine accuracy/systematic risks for LISA predictions, not self-reference. No exhibited reduction of a prediction to its own input is present, so the circularity score is 0.

Axiom & Free-Parameter Ledger

2 free parameters · 4 axioms · 0 invented entities

The central claims rest on the fidelity of the IMRPhenomXHM waveform model and the LISA response/noise model; no new entities are introduced. Free parameters are limited to chosen injection/analysis conventions.

free parameters (2)
  • Representative extrinsic parameters (δt, φ, λ_L, β_L, Ψ_L) = (0, 0.2, 1.8, π/6, 1.2)
    Fixed by hand for all injections; quantitative critical-redshift numbers and sky-octant hierarchy depend on this choice, though the qualitative conclusions are argued to be robust.
  • Critical redshift criterion (bias > 2σ statistical error) = 2σ
    Chosen threshold for defining 'significantly biased'; determines the z<2.5 event-fraction statement in the conclusions.
axioms (4)
  • standard math Wilks' theorem / chi-square distribution for log-likelihood differences
    Used in Eq. (2) to set the threshold ΔlnL < 9.84 for p=0.95, k=11; authors note it applies only to Gaussian likelihoods and caution about its use for very massive systems.
  • domain assumption LISA sensitivity is described by SciRDv1 PSD and TDI A/E/T channels
    Used throughout for inner products and SNR; affects all quantitative bias estimates.
  • domain assumption IMRPhenomXHM (version 122019) faithfully reproduces the true higher-mode content
    The entire systematic-bias calculation injects/recover with this model; authors acknowledge calibration uncertainty at high spins and the (3,3) artifact in Appendix B.2.
  • domain assumption Binaries are nonprecessing, quasicircular, spin-aligned
    State in intro/conclusions; precession and eccentricity are deferred to future work, so results do not cover those systems.

pith-pipeline@v1.3.0-alltime-deepseek · 26653 in / 12105 out tokens · 94589 ms · 2026-08-03T03:02:59.415630+00:00 · methodology

0 comments
read the original abstract

The Laser Interferometer Space Antenna (LISA) will observe massive black hole binaries (MBHBs) with astoundingly high signal-to-noise ratio, leaving parameter estimation with these signals susceptible to seemingly small waveform errors. Of particular concern for MBHBs are errors due to neglected higher-order modes. We extend Yi et al. [arXiv:2502.12237] to examine errors due to neglected higher-order modes for MBHBs with nonzero (aligned) progenitor spins and total mass up to $10^8\,M_\odot$. For these very massive systems, there can be regions of parameter space in which the $(\ell, |m|)=(2,\,2)$ modes are no longer dominant with respect to higher-order ones. We find that the extent of systematic bias can change significantly when varying the progenitor spins of the binary. We also find that for the heaviest, and therefore shortest, MBHB signals, slight systematic errors can cause severe misinference of the sky localization parameters. We propose an improved likelihood optimization scheme with respect to previous work as a way to predict these effects in a computationally efficient manner.

Figures

Figures reproduced from arXiv: 2602.09088 by Digvijay Wadekar, Emanuele Berti, Francesco Iacovelli, Nicol\'as Yunes, Rohit S. Chandramouli, Sophia Yi, Sylvain Marsat.

Figure 1
Figure 1. Figure 1: FIG. 1. SNR vs. detector-frame total mass for a few values [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. Hierarchy of SNR contribution by different angular harmonics as a function of total mass ( [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3. Hierarchy of SNR contribution by different angular [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 5
Figure 5. Figure 5: FIG. 5. Comparison of the systematic biases on two MBHB parameters with 4 modes [PITH_FULL_IMAGE:figures/full_fig_p007_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: FIG. 6. Same as Fig [PITH_FULL_IMAGE:figures/full_fig_p009_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: FIG. 7. Critical redshift at which the bias on intrinsic parameters due to neglecting the (3, 2) mode becomes greater than the [PITH_FULL_IMAGE:figures/full_fig_p010_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: FIG. 8. Same as Fig [PITH_FULL_IMAGE:figures/full_fig_p013_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: FIG. 9. The waveform error due to a neglected higher-order mode ((3, 2) in this case) can cause mis-localization of an MBHB. [PITH_FULL_IMAGE:figures/full_fig_p014_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: FIG. 10. Residual SNR [PITH_FULL_IMAGE:figures/full_fig_p014_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: FIG. 11. Histogram and scatter plot of the maximum log [PITH_FULL_IMAGE:figures/full_fig_p015_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: FIG. 12. Fractional SNR squared contributed by each harmonic, as a function of total mass and mass ratio. The inclination [PITH_FULL_IMAGE:figures/full_fig_p016_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: FIG. 13. Same as Fig [PITH_FULL_IMAGE:figures/full_fig_p017_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: FIG. 14. In the top panel, we show the [PITH_FULL_IMAGE:figures/full_fig_p018_14.png] view at source ↗

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. Improving low-latency multi-messenger follow-up of neutron star-black hole mergers with mode-by-mode filtering

    gr-qc 2026-06 unverdicted novelty 6.0

    Mode-by-mode filtering of higher-order modes enables low-latency marginalization over mode information in NSBH gravitational-wave signals, tightening constraints on distance, inclination, and secondary mass.

Reference graph

Works this paper leans on

59 extracted references · 45 linked inside Pith · cited by 1 Pith paper

  1. [1]

    downward

    Dual annealing The multidimensional, many-peaked nature of GW like- lihoods leaves simple optimization techniques susceptible to finding incorrect, local maxima. Some multimodalities, such as those due to near-degeneracies in the sky localiza- tion, are anticipated ahead of time given our knowledge of the detector response, and can therefore be addressed ...

  2. [2]

    antisymmet- ric spin

    Reparametrization We find that in instances where the optimizer can get stuck around local maxima, it is also useful to reparametrize the problem. This generally speeds up convergence and improves the robustness of our maxi- mization procedure. Similarly to Ref. [33], for the intrin- sic parameters, we use a basis inspired by the waveform fitting literatu...

  3. [3]

    NM” for Nelder-Mead algorithm only, as in Paper 1) with the improved likelihood optimization introduced in Sec. IIIB (“DA

    Using Fisher-informed priors We also find that the optimization can be sped up significantly if we restrict the parameter space based on the approximate width of the posteriors determined via Fisher analysis. To accommodate the fact that there can be very severe bias due to waveform inaccuracy, we still generally allow for exploration up to300σθ in either...

    2000

  4. [4]

    Amaro Seoaneet al.(LISA), (2017), arXiv:1702.00786 [astro-ph.IM]

    P. Amaro Seoaneet al.(LISA), (2017), arXiv:1702.00786 [astro-ph.IM]

    Pith/arXiv arXiv 2017

  5. [5]

    Colpiet al.(LISA), (2024), arXiv:2402.07571 [astro- ph.CO]

    M. Colpiet al.(LISA), (2024), arXiv:2402.07571 [astro- ph.CO]

    Pith/arXiv arXiv 2024

  6. [6]

    K. G. Arunet al.(LISA), Living Rev. Rel.25, 4 (2022), arXiv:2205.01597 [gr-qc]

    Pith/arXiv arXiv 2022

  7. [7]

    Afshordiet al.(LISA Consortium Waveform Working Group), Living Rev

    N. Afshordiet al.(LISA Consortium Waveform Working Group), Living Rev. Rel.28, 9 (2025), arXiv:2311.01300 [gr-qc]

    Pith/arXiv arXiv 2025

  8. [8]

    Kleinet al., Phys

    A. Kleinet al., Phys. Rev. D93, 024003 (2016), arXiv:1511.05581 [gr-qc]

    Pith/arXiv arXiv 2016

  9. [9]

    Rel.26, 5 (2023), arXiv:2204.05434 [astro-ph.CO]

    P.Auclairet al.(LISACosmologyWorkingGroup),Living Rev. Rel.26, 5 (2023), arXiv:2204.05434 [astro-ph.CO]

    Pith/arXiv arXiv 2023

  10. [10]

    Amaro Seoaneet al.(LISA), Living Rev

    P. Amaro Seoaneet al.(LISA), Living Rev. Rel.26, 2 (2023), arXiv:2203.06016 [gr-qc]

    Pith/arXiv arXiv 2023

  11. [11]

    Cutler and M

    C. Cutler and M. Vallisneri, Phys. Rev. D76, 104018 (2007), arXiv:0707.2982 [gr-qc]

    Pith/arXiv arXiv 2007

  12. [12]

    Lindblom, B

    L. Lindblom, B. J. Owen, and D. A. Brown, Phys. Rev. D78, 124020 (2008), arXiv:0809.3844 [gr-qc]

    Pith/arXiv arXiv 2008

  13. [13]

    Vallisneri, Phys

    M. Vallisneri, Phys. Rev. D86, 082001 (2012), arXiv:1207.4759 [gr-qc]

    Pith/arXiv arXiv 2012

  14. [14]

    J. R. Gair, M. Vallisneri, S. L. Larson, and J. G. Baker, Living Rev. Rel.16, 7 (2013), arXiv:1212.5575 [gr-qc]

    Pith/arXiv arXiv 2013

  15. [15]

    Yunes, X

    N. Yunes, X. Siemens, and K. Yagi, Living Rev. Rel.28, 3 (2025)

    2025

  16. [16]

    Pitte, Q

    C. Pitte, Q. Baghi, S. Marsat, M. Besançon, and A. Petiteau, Phys. Rev. D108, 044053 (2023), arXiv:2304.03142 [gr-qc]

    Pith/arXiv arXiv 2023

  17. [17]

    S. Yi, F. Iacovelli, S. Marsat, D. Wadekar, and E. Berti, Phys. Rev. D112, 124063 (2025), arXiv:2502.12237 [gr- qc]

    arXiv 2025

  18. [18]

    T.A.Prince, M.Tinto, S.L.Larson,andJ.W.Armstrong, Phys. Rev. D66, 122002 (2002), arXiv:gr-qc/0209039

    Pith/arXiv arXiv 2002

  19. [19]

    Babak, A

    S. Babak, A. Petiteau, and M. Hewitson, (2021), arXiv:2108.01167 [astro-ph.IM]

    Pith/arXiv arXiv 2021

  20. [20]

    García-Quirós, M

    C. García-Quirós, M. Colleoni, S. Husa, H. Estel- lés, G. Pratten, A. Ramos-Buades, M. Mateu-Lucena, and R. Jaume, Phys. Rev. D102, 064002 (2020), arXiv:2001.10914 [gr-qc]

    Pith/arXiv arXiv 2020

  21. [21]

    Pratten, S

    G. Pratten, S. Husa, C. Garcia-Quiros, M. Colleoni, A. Ramos-Buades, H. Estelles, and R. Jaume, Phys. Rev. D102, 064001 (2020), arXiv:2001.11412 [gr-qc]

    Pith/arXiv arXiv 2020

  22. [22]

    J. A. Nelder and R. Mead, Comput. J.7, 308 (1965)

    1965

  23. [23]

    Marsat, J

    S. Marsat, J. G. Baker, and T. Dal Canton, Phys. Rev. D103, 083011 (2021), arXiv:2003.00357 [gr-qc]

    Pith/arXiv arXiv 2021

  24. [24]

    W. D. Vousden, W. M. Farr, and I. Mandel, Mon. Not. Roy. Astron. Soc.455, 1919 (2016), arXiv:1501.05823

    Pith/arXiv arXiv 1919

  25. [25]

    Foreman-Mackey, D

    D. Foreman-Mackey, D. W. Hogg, D. Lang, and J. Goodman, Publ. Astron. Soc. Pac.125, 306 (2013), arXiv:1202.3665 [astro-ph.IM]

    Pith/arXiv arXiv 2013

  26. [26]

    S. S. Wilks, The Annals of Mathematical Statistics9, 60 (1938)

    1938

  27. [27]

    S. T. McWilliams, B. J. Kelly, and J. G. Baker, Phys. Rev. D82, 024014 (2010), arXiv:1004.0961 [gr-qc]

    Pith/arXiv arXiv 2010

  28. [28]

    Baird, S

    E. Baird, S. Fairhurst, M. Hannam, and P. Murphy, Phys. Rev. D87, 024035 (2013), arXiv:1211.0546 [gr-qc]

    Pith/arXiv arXiv 2013

  29. [29]

    Toubiana and J

    A. Toubiana and J. R. Gair, (2024), arXiv:2401.06845 [gr-qc]

    Pith/arXiv arXiv 2024

  30. [30]

    Baibhav, E

    V. Baibhav, E. Berti, and V. Cardoso, Phys. Rev. D101, 084053 (2020), arXiv:2001.10011 [gr-qc]

    Pith/arXiv arXiv 2020

  31. [31]

    Virtanen, R

    P. Virtanen, R. Gommers, T. E. Oliphant,et al., Nat. Methods17, 261 (2020)

    2020

  32. [32]

    Xiang, D

    Y. Xiang, D. Y. Sun, W. Fan, and X. G. Gong, Phys. Lett. A233, 216 (1997)

    1997

  33. [33]

    Tsallis, J

    C. Tsallis, J. Statist. Phys.52, 479 (1988)

    1988

  34. [34]

    Tsallis and D

    C. Tsallis and D. A. Stariolo, Physica A Statistical Me- chanics and its Applications233, 395 (1996), arXiv:cond- mat/9501047 [cond-mat]

    arXiv 1996

  35. [35]

    Xiang and X

    Y. Xiang and X. G. Gong, Phys. Rev. E62, 4473 (2000)

    2000

  36. [36]

    Roulet, S

    J. Roulet, S. Olsen, J. Mushkin, T. Islam, T. Venumadhav, B. Zackay, and M. Zaldarriaga, Phys. Rev. D106, 123015 (2022), arXiv:2207.03508 [gr-qc]

    Pith/arXiv arXiv 2022

  37. [37]

    K. K. Y. Ng, S. Vitale, A. Zimmerman, K. Chatziioannou, D. Gerosa, and C.-J. Haster, Phys. Rev. D98, 083007 (2018), arXiv:1805.03046 [gr-qc]

    Pith/arXiv arXiv 2018

  38. [38]

    Varma, D

    V. Varma, D. Gerosa, L. C. Stein, F. Hébert, and H. Zhang, Phys. Rev. Lett.122, 011101 (2019), arXiv:1809.09125 [gr-qc]

    Pith/arXiv arXiv 2019

  39. [39]

    Varma, S

    V. Varma, S. E. Field, M. A. Scheel, J. Blackman, L. E. Kidder, and H. P. Pfeiffer, Phys. Rev. D99, 064045 (2019), arXiv:1812.07865 [gr-qc]

    Pith/arXiv arXiv 2019

  40. [40]

    Fairhurst and P

    S. Fairhurst and P. Brady, Class. Quant. Grav.25, 105002 (2008), arXiv:0707.2410 [gr-qc]

    Pith/arXiv arXiv 2008

  41. [41]

    Berti, V

    E. Berti, V. Cardoso, J. A. Gonzalez, U. Sperhake, and B. Bruegmann, Class. Quant. Grav.25, 114035 (2008), arXiv:0711.1097 [gr-qc]. 20

    Pith/arXiv arXiv 2008

  42. [42]

    C. K. Mishra, A. Kela, K. G. Arun, and G. Faye, Phys. Rev. D93, 084054 (2016), arXiv:1601.05588 [gr-qc]

    Pith/arXiv arXiv 2016

  43. [43]

    London, S

    L. London, S. Khan, E. Fauchon-Jones, C. García, M. Hannam, S. Husa, X. Jiménez-Forteza, C. Kalaghatgi, F. Ohme, and F. Pannarale, Phys. Rev. Lett.120, 161102 (2018), arXiv:1708.00404 [gr-qc]

    Pith/arXiv arXiv 2018

  44. [44]

    Pompiliet al., Phys

    L. Pompiliet al., Phys. Rev. D108, 124035 (2023), arXiv:2303.18039 [gr-qc]

    Pith/arXiv arXiv 2023

  45. [45]

    C. B. Owen, C.-J. Haster, S. Perkins, N. J. Cor- nish, and N. Yunes, Phys. Rev. D108, 044018 (2023), arXiv:2301.11941 [gr-qc]

    Pith/arXiv arXiv 2023

  46. [46]

    A. Jan, D. Ferguson, J. Lange, D. Shoemaker, and A. Zimmerman, Phys. Rev. D110, 024023 (2024), arXiv:2312.10241 [gr-qc]

    Pith/arXiv arXiv 2024

  47. [47]

    Reitzeet al., Bull

    D. Reitzeet al., Bull. Am. Astron. Soc.51, 035 (2019), arXiv:1907.04833 [astro-ph.IM]

    Pith/arXiv arXiv 2019

  48. [48]

    Mangiagli, C

    A. Mangiagli, C. Caprini, M. Volonteri, S. Marsat, S. Ver- gani, N. Tamanini, and H. Inchauspé, Phys. Rev. D106, 103017 (2022), arXiv:2207.10678 [astro-ph.HE]

    Pith/arXiv arXiv 2022

  49. [49]

    Marsat and J

    S. Marsat and J. G. Baker, (2018), arXiv:1806.10734 [gr-qc]

    Pith/arXiv arXiv 2018

  50. [50]

    van der Steen, H

    T. van der Steen, H. Inchauspé, T. Hertog, and A. Hees, (2025), arXiv:2509.10038 [gr-qc]

    Pith/arXiv arXiv 2025

  51. [51]

    Marsat, in preparation

    S. Marsat, in preparation

  52. [52]

    Cornish, L

    N. Cornish, L. Sampson, N. Yunes, and F. Pretorius, Phys. Rev. D84, 062003 (2011), arXiv:1105.2088 [gr-qc]

    Pith/arXiv arXiv 2011

  53. [53]

    R. S. Chandramouli, K. Prokup, E. Berti, and N. Yunes, Phys. Rev. D111, 044026 (2025), arXiv:2410.06254 [gr- qc]

    Pith/arXiv arXiv 2025

  54. [54]

    Barausse, K

    E. Barausse, K. Dey, M. Crisostomi, A. Panayada, S. Marsat, and S. Basak, Phys. Rev. D108, 103034 (2023), arXiv:2307.12245 [astro-ph.GA]

    Pith/arXiv arXiv 2023

  55. [55]

    Littenberg, N

    T. Littenberg, N. Cornish, K. Lackeos, and T. Robson, Phys. Rev. D101, 123021 (2020), arXiv:2004.08464 [gr- qc]

    Pith/arXiv arXiv 2020

  56. [56]

    T. B. Littenberg and N. J. Cornish, Phys. Rev. D107, 063004 (2023), arXiv:2301.03673 [gr-qc]

    Pith/arXiv arXiv 2023

  57. [57]

    M. L. Katz, N. Karnesis, N. Korsakova, J. R. Gair, and N. Stergioulas, Phys. Rev. D111, 024060 (2025), arXiv:2405.04690 [gr-qc]

    Pith/arXiv arXiv 2025

  58. [58]

    S. Deng, S. Babak, M. Le Jeune, S. Marsat, É. Plagnol, and A. Sartirana, Phys. Rev. D111, 103014 (2025), arXiv:2501.10277 [gr-qc]

    Pith/arXiv arXiv 2025

  59. [59]

    Khukhlaev and S

    S. Khukhlaev and S. Babak, Phys. Rev. D112, 123046 (2025), arXiv:2509.20062 [gr-qc]

    arXiv 2025