Pith. sign in

REVIEW 3 major objections 6 minor 2 cited by

High-frequency continuous gravitational waves searched in LIGO O3 public data with Einstein@Home

T0 review · 3 major / 6 minor · reviewed 2026-08-05 · deepseek-v4-flash

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

Pith's one-line read The paper reports the most sensitive all-sky search to date for high-frequency continuous gravitational waves, finds no signal, and sets upper limits reaching h0 = 1.32e-25 at 800 Hz.

desk verdict A genuinely more sensitive all-sky CW search that recovers injection 7, whose weakest point is the documented but incomplete disturbance screening; still worth refereeing carefully. read the letter →

arxiv 2508.20073 v2 pith:JWBISLAW submitted 2025-08-27 gr-qc

Brian McGloughlin , Benjamin Steltner , Jasper Martins , Maria Alessandra Papa , Heinz-Bernd Eggenstein , Jing Ming , Bernd Machenschalk , Reinhard Prix
show 1 more author
Maximillian Bensch
This is my paper
classification gr-qc PACS 04.80.Nn95.85.Sz
keywords continuousgravitationalwavesall-skysearchneutronstarellipticityr-modesupperlimitsLIGOO3semi-coherentvolunteercomputing
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

This paper searches for nearly monochromatic gravitational waves from spinning neutron stars in the 800–1686 Hz band using six months of public LIGO data, processed on volunteer computers. It finds no astrophysical signal, and it turns that absence into the tightest upper limits yet set in this frequency and spin-down range: at 800 Hz, gravitational-wave amplitudes above 1.32e-25 are excluded at 90% confidence. If the result holds, isolated neutron stars spinning faster than 2.5 ms with ellipticities above about 2e-8 at 100 pc do not exist, and r-mode amplitudes above 7e-7 at 100 pc are ruled out for stars spinning above 400 Hz. The search also recovers both hardware-injected signals in its range, including the first recovery of the 1220 Hz injection, which exercises the full detection chain. The limits depend on excluding a small set of noise-disturbed frequency bands, and the paper is explicit about which bands are excluded and why.

What carries the argument

The search is a five-stage hierarchy. Stage 0 uses the semi-coherent Global Correlation Transform with the multi-detector F-statistic on 60-hour segments over 1.4e18 templates; its top-lists are visually inspected and clustered into about 13 million seeds. Stage 1 refines each seed with a finer grid and the line-robust betaS/GLtL statistic, applying veto thresholds. Stages 2 and 3 switch to a Bayesian nested-sampling follow-up that estimates evidences over 120-hour and then fully coherent baselines. The sensitivity target is a detection depth of about 56 per sqrt(Hz) at 90% efficiency, matching the previous lower-frequency search; the actual setup achieves this by accepting a slightly lower

What would settle it

Run the full Stage 0–3 hierarchy on the non-excluded bands with software-injected signals drawn uniformly from the target sensitivity-depth population, and measure the fraction recovered above the Stage 1 veto thresholds; if fewer than 90% of injections at h0 equal to the quoted limit are recovered, the stated upper limits overestimate the search's sensitivity.

Watch

Extended reading notes

Core claim

In the parameter space 800.0–1686.0 Hz with spin-down between −2.7e-9 and 0.2e-9 Hz/s, after sifting 1.4e18 templates down to 13 million candidates and following them through a grid-based then Bayesian hierarchy, no candidate survives as an astrophysical signal. The paper therefore reports a non-detection and sets 90% confidence upper limits on the intrinsic gravitational-wave amplitude, the most stringent being h0 = 1.32e-25 at 800 Hz. Under the triaxial-ellipsoid model for a rotating neutron star, these limits exclude ellipticities epsilon >= 1.96e-8 (d/100 pc) for stars spinning faster than 2.5 ms, and under the r-mode model they exclude alpha >= 7e-7 (d/100 pc) for stars spinning faster

Load-bearing premise

The quoted upper limits assume that every noise-disturbed frequency band was correctly flagged and excluded, and that line-cleaning removed only noise, never astrophysical signal; the paper itself records one disturbed violin-mode band that evaded the initial visual check and 79 bands where signals were removed together with lines.

Editorial extensions

If this is right

  • No isolated neutron star in the searched band, spin-down range, and distance is deformed strongly enough to radiate above the quoted amplitudes; the ellipticity exclusions for fast rotators hold within the model assumptions.
  • The r-mode upper limits independently constrain a second emission mechanism, ruling out r-mode amplitudes above 7e-7 (d/100 pc) for stars spinning above 400 Hz.
  • First recovery of hardware injection 7 validates that a real continuous-wave signal in the high-frequency band would be found and its parameters pinned down by this pipeline.
  • The search shows that volunteer-computing power can cover the computationally expensive high-frequency sky at the same sensitivity depth as lower-frequency searches.
  • Because 79 half-Hz bands had signals removed together with lines, and several other bands are excluded, the upper limits do not constrain those specific frequencies; the accompanying supplementary list is needed to interpret the limits.

Reading between the lines

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

  • The clear separation between candidate and injected-signal evidence at the coherent stage suggests the same hierarchy could be run with an automatic disturbance classifier in place of visual inspection; a comparison of automatically flagged bands with the paper's manual exclusions would test whether any disturbed bands were missed.
  • The distance-scaled exclusion, epsilon >= 1.96e-8 (d/100 pc), implies that any known nearby isolated neutron star closer than about 100 pc with a deformation above this level would have been detected; a targeted re-analysis of the nearest candidates in this frequency band would test the same physics at lower computational cost.
  • Applied to the next, longer observation run, the same pipeline should lower the amplitude limits by roughly the square-root of the coherent time ratio, provided the noise disturbance identification keeps pace.
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

3 major / 6 minor

Summary. The paper reports an all-sky search for continuous gravitational waves in LIGO O3a public data, covering frequencies 800.0–1686.0 Hz and spin-downs between -2.7e-9 and 0.2e-9 Hz/s. The search uses a four-stage hierarchy: a semi-coherent GCT Stage 0 on Einstein@Home, a finer-grid Stage 1, and two Bayesian follow-up stages (Stage 2 semi-coherent, Stage 3 fully coherent). The authors report no gravitational-wave candidates, recover both hardware injections in the search range (including the first reported recovery of hardware injection 7), and set 90% upper limits on h0, with the most stringent limit h0 = 1.32e-25 at 800 Hz. They translate these limits into constraints on neutron-star ellipticity and r-mode amplitude. Approximately 6% of the searched frequency range is excluded from the upper-limit statement because of identified disturbances or line-cleaning side effects.

Significance. If the upper limits are valid, this is the most sensitive all-sky search in this high-frequency parameter space, extending previous Einstein@Home results and probing astrophysically interesting ellipticities and r-modes. The search is carefully calibrated: injection-recovery studies set and validate the sensitivity depth, the pipeline recovers both hardware injections, and the Bayesian follow-up is a novel and computationally efficient approach. The availability of a supplementary excluded-band list is a strength. However, the central upper-limit claims rest on the completeness of the disturbance-identification procedure, and the paper itself documents a case where that procedure failed. This makes the validity of the quoted limits, and the 'most sensitive search' claim, conditional on an assumption that is not fully demonstrated.

major comments (3)
  1. [Section 5.3 and Section 6.2] The paper explicitly documents that the Stage 0/1 disturbance screening is incomplete: the 1500–1501 Hz violin-mode band 'was not flagged' by visual inspection at Stage 0, and Stage 1 results were 'largely unremarkable'; it was recognized only at Stage 3. This is direct evidence that a non-Gaussian disturbance can survive the screening that determines which bands are used for upper limits. Since upper limits are set in half-Hz bands using detection efficiency measured from injections, an unflagged disturbance in any remaining band would bias the measured efficiency high and the h0 upper limit low. The excluded-band list in the supplementary materials is therefore load-bearing, but its completeness is not established. The authors should either (a) provide a systematic validation that every band used for upper limits has detection efficiency consistent with the clean-noise model (e.g., usi
  2. [Section 6.2, item 4] The exclusion of 79 half-Hz bands because line-cleaning removes signals together with lines is a reasonable and honest limitation, but the paper does not quantify how these exclusions affect the headline sensitivity comparison in Figure 7. The 'most sensitive search' claim is made without stating the fraction of the parameter space that is actually usable for upper limits (excluded bands are approximately 6% of the frequency range). A reader comparing this search to others (e.g., Dergachev & Papa 2025a) cannot see whether the comparison accounts for the different excluded regions. The authors should state the usable parameter-space fraction and, if appropriate, note that the sensitivity comparison is made on the common usable region. This is not a fatal flaw, but it is necessary for a fair assessment of the central claim.
  3. [Section 6.2 / Methodology of upper limits] The upper-limit calculation is described only by reference to Steltner et al. (2023). For a paper whose main quantitative results are these limits, the description is too terse. In particular, it is not stated how detection efficiency is measured in each half-Hz band: whether the injection-recovery population is drawn separately for every band, how the measured efficiency is combined with the loudest-candidate distribution to produce the 90% upper limit, and how the excluded sub-bands enter the calculation. Without this detail, the reader cannot judge whether the documented disturbance-screening failure could affect a specific band's limit (e.g., the 800 Hz best limit). The authors should include a concise description or explicitly cite the exact equations in Steltner et al. (2023) that are used.
minor comments (6)
  1. [Abstract] Typo: 'neutron stars stars spinning faster than 400 Hz' should be 'neutron stars spinning faster than 400 Hz'.
  2. [Section 1] Typo: 'surrounding back holes' should be 'surrounding black holes'.
  3. [Section 3.1] 'remove noise that would degrade the quality of the search results, lines and in the frequency domain' is missing a word; likely 'lines in the frequency domain'.
  4. [Figure 2] The y-axis label contains a typo: 'efficency' should be 'efficiency'.
  5. [Section 6.2] The sentence 'The upper limit set in a half-Hz band may not hold in the whole band' is slightly confusing; consider rephrasing to 'The upper limit is not claimed for the entire half-Hz band when sub-bands are excluded'.
  6. [Table 1] In the Bayesian-stages row header, 'Jeffreys'' appears without a second apostrophe; likely should be 'Jeffreys' prior'.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: upper limits come from injection-recovery efficiency, benchmarked against external hardware injections; the missed violin-mode band is a data-quality limitation, not a circular step.

full rationale

The paper's central claims—the h0 upper limits and the derived ellipticity/r-mode exclusions—are produced by injection-and-recovery studies: synthetic signals are added to the data, the pipeline's detection efficiency is measured as a function of sensitivity depth, and the upper limit is the amplitude at which 90% efficiency is achieved. This is a standard empirical calibration, not a definitional identity: the injected signals are not the target result, and the hardware injections (IDs 1 and 7) are external benchmarks added at the detector level and recovered by the search, providing independent validation. The Stage 1 veto thresholds are calibrated on an injected reference population, but this calibration is a sensitivity-tuning step, not a disguised prediction; the same thresholds are then applied to search candidates, and the hardware injections are recovered without being used in the fitting. The Bayesian follow-up method is cited from Martins et al. (2025), which shares authors with this paper, but its recovery performance is demonstrated in-paper on the target signal population and on hardware injections, so the self-citation is not load-bearing. The paper is unusually explicit about its limitations: Section 5.3 documents that the 1500–1501 Hz violin-mode band was missed by Stage 0 visual inspection and had to be excluded at a later stage, and Section 6.2 lists 79 half-Hz bands excluded because line cleaning removes signals. These are honest correctness risks (an unflagged quiet line could make the quoted limits too strong), but they are not circularity: they reduce the scope of the upper-limit statement rather than redefining the output as an input. No equation in the derivation chain is equivalent to its own inputs by construction, and the hardware-injection recovery anchors the search to external reality.

Assumptions & free parameters 6 free parameters · 6 assumptions · 0 invented entities

The search rests on standard continuous-wave signal models and detection statistics, plus domain assumptions about detector noise after cleaning and about the mapping from amplitude to neutron star properties. No new physical entities are introduced. All free parameters are search-design choices (grids, coherence times, veto thresholds) tuned to a sensitivity target inherited from Steltner et al. (2023); none is fitted to define the upper limits, which are calibrated by injection-recovery against hardware injections.

free parameters (6)
  • msky sky-grid coarseness = 0.008 (Stage 0), 0.002 (Stage 1)
    Tunable free parameter controlling sky-grid coarseness (Eq. 6, Section 3.3.1); chosen to match the sensitivity-depth target within the compute budget.
  • grid spacings delta_f, delta_fdot, gamma1 = 4/2 microHz, 1.5/0.9 x 1e-10 Hz/s, gamma1 = 450
    Template grid design choices (Table 1) tuned so Stage 0 reaches the target sensitivity depth; they set resolution, not the physical result.
  • coherence times Tcoh = 60 h (Stages 0-1), 120 h (Stage 2), full (Stage 3)
    Chosen from the runtime-versus-sensitivity trade-off (Figure 2) within the six-month Einstein@Home budget.
  • sensitivity depth target and bracket = D = [45, 65], goal about 56 1/sqrt(Hz)
    A priori target set to match Steltner et al. (2023); not fitted to this search's data but defines the search configuration.
  • Stage-1 veto thresholds = 2F_thr = 6.76, log10 beta_thr = 0
    Calibrated on the injected reference population (Eq. 9, Section 5.1): a tuning of the follow-up cut to the simulation, not to astrophysical data.
  • Bayesian stage settings = nlive 750/250, Jeffreys/GMM priors
    Nested-sampling configuration (Table 1) from the companion method paper; choices affect cost and accuracy, not the physical result.
assumptions (6)
  • standard math CW signal model: slowly spinning-down triaxial source with phase given by Eq. 4, depending only on f and fdot, nearly monochromatic over weeks
    Standard model of Jaranowski et al. (1998) and Cutler & Schutz (2005), invoked in Section 2; higher spin-down derivatives and free precession are excluded by design.
  • standard math Detection statistic 2F is chi-squared with 4Nseg degrees of freedom under Gaussian noise (Eq. 8)
    Frequentist F-statistic theory used for the Stage 0 efficiency estimate (Section 4.1).
  • domain assumption After line and glitch cleaning, detector noise in non-excluded bands is sufficiently Gaussian for the efficiency model
    The efficiency estimate and upper limits require this; Section 5.3 shows it fails in the 1500-1501 Hz violin-mode band, which the search recognized only late, so this assumption is fragile.
  • standard math GCT semi-coherent search approximates the fully-coherent statistic at negligible mismatch for the chosen grids
    Pletsch & Allen (2009) and Pletsch (2010), used in Section 3.2; the efficiency at the grid's average mismatch is part of the calibration.
  • domain assumption Line cleaning removes noise lines without removing astrophysical signals in bands where limits are quoted
    Section 6.2 (item 4) states that in 79 half-Hz bands the cleaning does remove signals, so the assumption holds only for the remaining bands; the paper excludes the contaminated bands from limits.
  • domain assumption Ellipticity and r-mode translations (Eqs. 11-12) with I = 1e38 kg m2 and the Owen (2010) r-mode model
    Standard mapping from h0 to astrophysical parameters in Section 6.2; the moment of inertia and emission model are taken from prior literature.

how reviews work

0 comments
Cite this review

Pith. "Pith review of High-frequency continuous gravitational waves searched in LIGO O3 public data with Einstein@Home." pith.science (2026). https://pith.science/paper/JWBISLAW

@misc{pith2026250820073,
  author       = {Pith},
  title        = {Pith review of: High-frequency continuous gravitational waves searched in LIGO O3 public data with Einstein@Home},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JWBISLAW}},
  note         = {Machine review of arXiv:2508.20073}
}
abstract

We search for nearly-monochromatic gravitational wave signals with frequencies $800.0~\textrm{Hz} \leq f \leq 1686.0~\textrm{Hz}$ and spin-down $-2.7\times10^{-9}~\textrm{Hz}\,\textrm{s}^{-1} \leq \dot f \leq 0.2\times 10^{-9}~\textrm{Hz}\,\textrm{s}^{-1}$. We use LIGO O3 public data from the Hanford and Livingston detectors and deploy this search on the Einstein@Home volunteer-computing project. This is the most sensitive search carried out to date in this parameter space. Our results are consistent with a non-detection. We set upper limits on the gravitational wave amplitude $h_{0}$ and translate these to upper limits on neutron star ellipticity and on r-mode amplitude. The most stringent upper limits are at $800~\textrm{Hz}$ with $h_{0} = 1.32\times10^{-25}$, at the $90\%$ confidence level. Searching in the high frequency bands allows us to probe astrophysically interesting ellipticities with our results excluding isolated neutron stars rotating faster than $2.5~\textrm{ms}$ with ellipticities $\epsilon \geq 1.96 \times 10^{-8}\left[\frac{d}{100~\textrm{pc}}\right]$ within a distance $d$ from Earth. Our results also exclude r-mode amplitudes $\alpha \geq 7 \times 10^{-7}\left[\frac{d}{100~\textrm{pc}}\right]$ for neutron stars stars spinning faster than 400 Hz.

Figures

Figures reproduced from arXiv: 2508.20073 by the authors.

Figure 1
Figure 1. 0 50 100 150 LHO SFTs LLO SFTs Stage 0 - 1 Stage 2 (Bayesian) Stage 3 (Bayesian) Days after 2019 Apr 1 [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 3
Figure 3. The number of Stage 0 candidates from every 50 mHz band. The outliers are discussed in Section 4.3. The relatively large amplitude of hardware injection 1 in the 848.90 Hz band saturates the results and produces fewer but larger clusters in this 50 mHz band. Visual Inspection: All 50 mHz results are visually in￾spected to identify disturbed bands Abbott et al. (2017). We do not set upper limits in these bands becaus… view at source ↗
Figure 4
Figure 4. shows the Stage 1 candidates from the search and from the target signal population. The distributions for the candidates and the reference population have not yet separated, and further follow-up is necessary. Based on the signal population results, we define a veto that identifies candidates that should pass on to the next stage: a candidate survives if    2F¯cand ≥ 2F¯thr = 6.76 log10 βˆcand S/GLtL ≥ log10 βˆth… view at source ↗
Figures from the paper (4 more)
Figure 5
Figure 5. Figure 5: The results of the Bayesian searches in Stage 2 and 3 for the signal candidates and the reference population of test signals. The upper plot shows the evidences at Stage 2, Z 2 . For Stage 3, we show the integrated evidences Z 3 and their relative change from Stage 2, …
Figure 6
Figure 6. Figure 6: The posterior distribution for hardware injection 7 after search Stage 3. The signal parameters are given at our reference time 1246070525.0 (GPS time). which were obtained using both LIGO O3a and O3b data, corresponding to an approximately twice as long time-baseline.…
Figure 7
Figure 7. Figure 7: Smallest gravitational wave amplitude h0 that our results exclude at 90% confidence. We compare to other all-sky searches in LIGO O3 data (Abbott et al. 2022; Dergachev & Papa 2025a). For Dergachev & Papa (2025a) two curves are shown, as they set best and worst case up…
Figure 9
Figure 9. Figure 9: shows the upper limits on the r-mode amplitude α for different distances d. 800 1000 1200 1400 1600 Signal frequency [Hz] 10−8 10−7 10−6 10−5 10−4 Source ellipticity ε max ε 4 kpc 1 kpc 100 pc [PITH_FULL_IMAGE:figures/full_fig_p009_9.png]

Discussion (0). Sign in to comment.

Forward citations

Cited by 2 Pith papers

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

  1. First Constraints on the Ellipticities of Self-Interacting Fermionic Dark Matter Admixed Neutron Stars from Continuous Gravitational-Wave Searches

    astro-ph.CO 2026-06 unverdicted novelty 7.0 of 10

    Using LIGO O3 continuous-wave search data, the authors place the first constraints on ellipticities of self-interacting fermionic dark matter admixed neutron stars and exclude regions of the DM parameter space for mas...

  2. Search for continuous gravitational waves from the pulsar J0435+3233

    gr-qc 2026-07 accept novelty 6.0 of 10

    A LIGO O4a search for continuous gravitational waves from millisecond pulsar J0435+3233 finds no signal, setting h0<5.8×10^-27 at 95% confidence and an ellipticity limit of 1.6×10^-8.

Reference graph

Works this paper leans on

42 extracted references · 12 canonical work pages · cited by 2 Pith papers

  1. [1]

    P., et al

    Abbott, B. P., et al. 2017, Phys. Rev. D, 96, 122004, 10.1103/PhysRevD.96.122004

    work page doi:10.1103/physrevd.96.122004 2017

  2. [2]

    2022, Phys

    Abbott, R., et al. 2022, Phys. Rev. D, 106, 102008, 10.1103/PhysRevD.106.102008

    doi:10.1103/physrevd.106.102008 2022

  3. [3]

    Anderson , D. P. 2004, in Proceedings of the Fifth IEEE/ACM International Workshop on Grid Computing (GRID04), 4--10

    work page 2004

  4. [4]

    P., Christensen , C., & Allen , B

    Anderson , D. P., Christensen , C., & Allen , B. 2006, in Proceedings of the 2006 ACM/IEEE conference on Supercomputing, 126--136

    work page 2006

  5. [5]

    2019, The Astrophysical Journal Supplement Series, 241, 27, 10.3847/1538-4365/ab06fc

    Ashton, G., et al. 2019, The Astrophysical Journal Supplement Series, 241, 27, 10.3847/1538-4365/ab06fc

    doi:10.3847/1538-4365/ab06fc 2019

  6. [6]

    A., & Prix, R

    Behnke, B., Papa, M. A., & Prix, R. 2015, Phys. Rev. D, 91, 064007, 10.1103/PhysRevD.91.064007

    doi:10.1103/physrevd.91.064007 2015

  7. [7]

    2020, The permanent ellipticity of the neutron star in PSR J1023+0038 , 10.1093/mnras/staa2304

    Bhattacharyya, S. 2020, The permanent ellipticity of the neutron star in PSR J1023+0038 , 10.1093/mnras/staa2304

    work page doi:10.1093/mnras/staa2304 2020

  8. [8]

    2017, Phys

    Biwer, C., et al. 2017, Phys. Rev. D, 95, 062002, 10.1103/PhysRevD.95.062002

    doi:10.1103/physrevd.95.062002 2017

Show all 42 references
  1. [9]

    R., & Creighton, T

    Brady, P. R., & Creighton, T. 2000, Phys. Rev. D, 61, 082001, 10.1103/PhysRevD.61.082001

    2000 doi

  2. [10]

    R., Creighton, T., Cutler, C., & Schutz, B

    Brady, P. R., Creighton, T., Cutler, C., & Schutz, B. F. 1998, Phys. Rev. D, 57, 2101, 10.1103/PhysRevD.57.2101

    1998 doi

  3. [11]

    2025, Phys

    Capote, E., et al. 2025, Phys. Rev. D, 111, 062002, 10.1103/PhysRevD.111.062002

    2025 doi

  4. [12]

    2018, Phys

    Covas, P., et al. 2018, Phys. Rev. D, 97, 082002, 10.1103/PhysRevD.97.082002

    2018 doi

  5. [13]

    Cutler, C., & Schutz, B. F. 2005, Phys. Rev. D, 72, 063006, 10.1103/PhysRevD.72.063006

    2005 doi

  6. [14]

    J., Lundgren, A

    Davis, D., Massinger, T. J., Lundgren, A. P., et al. 2019, Class. Quant. Grav., 36, 055011, 10.1088/1361-6382/ab01c5

    2019 doi

  7. [15]

    2021, Class

    Davis, D., et al. 2021, Class. Quant. Grav., 38, 135014, 10.1088/1361-6382/abfd85

    2021 doi

  8. [16]

    Dergachev, V., & Papa, M. A. 2025 a , Phys. Rev. D, 112, 042005, 10.1103/psrl-y44w

    2025 doi

  9. [17]

    ---. 2025 b . 2507.12161

    2025 arXiv

  10. [18]

    2021, LIGO-T2100200-v2 : O3 Lines and Combs in Found in Self-Gated C01 Data, https://dcc.ligo.org/LIGO-T2100200/public

    Goetz, E., et al. 2021, LIGO-T2100200-v2 : O3 Lines and Combs in Found in Self-Gated C01 Data, https://dcc.ligo.org/LIGO-T2100200/public

    2021

  11. [19]

    Hessels, J. W. T., Ransom, S. M., Stairs, I. H., et al. 2006, Science, 311, 1901, 10.1126/science.1123430

    2006 doi

  12. [20]

    Jaranowski, P., Kr \' o lak, A., & Schutz, B. F. 1998, Physical Review D, 58, 10.1103/physrevd.58.063001

    1998 doi

  13. [21]

    1946, Proceedings of the Royal Society of London

    Jeffreys, H. 1946, Proceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences, 186, 453, 10.1098/rspa.1946.0056

    1946

  14. [22]

    K., & Owen, B

    Johnson-McDaniel, N. K., & Owen, B. J. 2013, Phys. Rev. D, 88, 044004, 10.1103/PhysRevD.88.044004

    2013 doi

  15. [23]

    2016, Phys

    Keitel, D. 2016, Phys. Rev. D, 93, 084024, 10.1103/PhysRevD.93.084024

    2016 doi

  16. [24]

    M., Papa, M

    Krishnan, B., Sintes, A. M., Papa, M. A., et al. 2004, Phys. Rev. D, 70, 082001, 10.1103/PhysRevD.70.082001

    2004 doi

  17. [25]

    C., & Virgo, P

    LIGO, S. C., & Virgo, P. 2022, https://www.gw-openscience.org/O3/O3April1_injection_parameters/

    2022

  18. [26]

    A., Steltner, B., Prix, R., & Vidal, P

    Martins, J., Papa, M. A., Steltner, B., Prix, R., & Vidal, P. B. 2025. 2508.18204

    2025

  19. [27]

    McGloughlin, B., Martins, J., Steltner, B., et al. 2025 a . 2508.16423

    2025 arXiv

  20. [28]

    McGloughlin, B., Steltner, B., Martins, J., et al. 2025 b , Supplemental materials to the paper High-frequency continuous gravitational waves searched in LIGO O3 public data with Einstein@Home , www.aei.mpg.de/continuouswaves/EaH-O3ASHi

    2025

  21. [29]

    A., & Horowitz, C

    Morales, J. A., & Horowitz, C. J. 2024, Phys. Rev. D, 110, 044016, 10.1103/PhysRevD.110.044016

    2024 doi

  22. [30]

    Owen, B. J. 2010, Phys. Rev. D, 82, 104002, 10.1103/PhysRevD.82.104002

    2010 doi

  23. [31]

    J., Lindblom, L., Cutler, C., et al

    Owen, B. J., Lindblom, L., Cutler, C., et al. 1998, Phys. Rev. D, 58, 084020, 10.1103/PhysRevD.58.084020

    1998 doi

  24. [32]

    A., Ming, J., et al

    Pagliaro, G., Papa, M. A., Ming, J., et al. 2023, The Astrophysical Journal, 952, 123, 10.3847/1538-4357/acd76f

    2023 doi

  25. [33]

    2011, Journal of machine learning research, 12, 2825

    Pedregosa, F., Varoquaux, G., Gramfort, A., et al. 2011, Journal of machine learning research, 12, 2825

    2011

  26. [34]

    Pletsch, H. J. 2010, Phys. Rev. D, 82, 042002, 10.1103/PhysRevD.82.042002

    2010 doi

  27. [35]

    J., & Allen, B

    Pletsch, H. J., & Allen, B. 2009, Phys. Rev. Lett., 103, 181102, 10.1103/PhysRevLett.103.181102

    2009 doi

  28. [36]

    2004, AIP Conference Proceedings, 735, 395, 10.1063/1.1835238

    Skilling, J. 2004, AIP Conference Proceedings, 735, 395, 10.1063/1.1835238

    2004 doi

  29. [37]

    Speagle, J. S. 2020, Monthly Notices of the Royal Astronomical Society, 493, 3132, 10.1093/mnras/staa278

    2020 doi

  30. [38]

    A., & Eggenstein, H.-B

    Steltner, B., Menne, T., Papa, M. A., & Eggenstein, H.-B. 2022 a , Phys. Rev. D , 106 , 104063 , 10.1103/PhysRevD.106.104063

    2022 doi

  31. [39]

    A., & Eggenstein, H.-B

    Steltner, B., Papa, M. A., & Eggenstein, H.-B. 2022 b , Phys. Rev. D, 105, 022005, 10.1103/PhysRevD.105.022005

    2022 doi

  32. [40]

    A., Eggenstein, H

    Steltner, B., Papa, M. A., Eggenstein, H. B., et al. 2023, Astrophys. J., 952, 55, 10.3847/1538-4357/acdad4

    2023 doi

  33. [41]

    Taylor , M. B. 2005, in Astronomical Society of the Pacific Conference Series, Vol. 347, Astronomical Data Analysis Software and Systems XIV, ed. P. Shopbell , M. Britton , & R. Ebert , 29

    2005

  34. [42]

    J., Baryakhtar, M., Papa, M

    Zhu, S. J., Baryakhtar, M., Papa, M. A., et al. 2020, Phys. Rev. D, 102, 063020, 10.1103/PhysRevD.102.063020

    2020 doi

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.