Pith. sign in

REVIEW 4 major objections 4 minor 70 references

Exploring the parameter space of hierarchical triple black hole systems

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

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

Pith's one-line read A seven-dimensional map of nearly 15 million hierarchical triple black hole systems shows which initial configurations merge within 14 Gyr, and a neural network predicts the outcome with 95% accuracy (99.7% at high confidence).

desk verdict A genuinely large, transparent 7D map and a fast surrogate for triple-BH merger outcomes, but the headline accuracy numbers are validated only against the same secular code that generated the labels, and the reported merger fractions inherit an adaptive sampling bias. read the letter →

arxiv 2506.22519 v1 pith:BZUQOTMM submitted 2025-06-26 astro-ph.HE astro-ph.GAastro-ph.SRgr-qc

M. Attia , Y. Sibony This is my paper
classification astro-ph.HEastro-ph.GAastro-ph.SRgr-qc
keywords blackholephysicsgravitationalwaveshierarchicaltriplesystemsvonZeipel-Lidov-KozaimechanismsecularapproximationMarkovchainMonteCarloneuralnetworkpredictioninitialseparationproblem
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper sets out to solve the 'initial separation' problem of gravitational-wave astrophysics: isolated black hole binaries must start extremely close to merge within a Hubble time, yet their stellar progenitors could not have been that close. The authors argue that hierarchical triples offer a way out, and they map which initial configurations of such triples actually merge within 14 Gyr. Using a secular orbit-averaged code with gravitational-wave losses and an adaptive MCMC sampler, they explore nearly 15 million systems across seven parameters and identify the merger-prone regions: asymmetric inner masses, moderately large inner separations where the von Zeipel-Lidov-Kozai (ZLK) mechanism is not quenched by relativistic precession, small outer separations, large outer eccentricities, and higher mutual inclination. They also train a neural network that predicts merger outcome with 95% accuracy (99.7% on high-confidence predictions), making fast population synthesis feasible. A sympathetic reader would take the central claim to be that the hierarchical triple channel can indeed resolve the initial separation problem for a well-characterized slice of parameter space.

What carries the argument

The engine of the argument is the modified JADE secular code: the Hamiltonian of the hierarchical triple is expanded in Legendre polynomials and truncated at hexadecapolar order, so the equations of motion are averaged over both orbits, and the standard gravitational-wave dissipation formulas are added to the inner binary. On top of that, an adaptive MCMC sampling scheme uses a k-nearest-neighbour score to concentrate simulations along the merger/nonmerger transition, refining the score function as new samples arrive. The von Zeipel-Lidov-Kozai mechanism (secular eccentricity oscillations of the inner binary driven by an inclined, distant third body) is the physical process that carries the merger channel, and the neural network is a trained surrogate that maps the seven initial parameters to merger probability.

What would settle it

Run the same 14 Gyr evolution for the roughly 500 uncertain-boundary systems (posterior merger probability around 0.5) with a direct N-body integrator; the paper reports 79% agreement there, so if the true agreement falls well below that, the mapped boundary would shift and the network's 95% accuracy would not transfer to an independent uniform sample. A second check is to retrain the network on a uniform grid rather than the MCMC-weighted sample and compare accuracy on the same test set.

Watch

Extended reading notes

Core claim

The central discovery is the first systematic seven-dimensional map of the merger boundary for hierarchical triple black holes, built from 14.4 million unique secular simulations. The boundary is not smooth: merger probability rises with mutual inclination but shows robust peaks near 54 and 64 degrees separated by a trough at 60 degrees, and the authors interpret the irregular structure as evidence of chaotic, possibly fractal, dynamics in the three-body problem. Within the surveyed box, merger-favorable systems have asymmetric inner binary masses (benefiting from octupolar ZLK excitation), inner separations large enough for ZLK oscillations to operate but small enough for GW emission to matter, outer separations near the stability limit, and outer eccentricities above about 0.6. The trained neural network reproduces these outcomes with an ROC AUC of 99%, and for the roughly 80% of predictions it makes with confidence above 0.9, accuracy reaches 99.7%.

Load-bearing premise

The whole map and the neural network inherit whatever error lives in the simplified orbit-averaged secular model, which agrees with full N-body integration on only 79% of the borderline systems it was specifically designed to sample.

Editorial extensions

If this is right

  • Systems with asymmetric inner binary masses, moderate-large inner separations, tight outer orbits, and outer eccentricities above about 0.6 are the ones most likely to merge within 14 Gyr, so population synthesis should weight these regions higher.
  • The neural network classifies a system in milliseconds, so full dynamical integration is only needed for the roughly 20% of low-confidence cases; this makes large-scale population synthesis of the hierarchical triple channel computationally tractable.
  • The four-way classification of nonmergers (GW+ZLK, GW-only, ZLK-only, neither) identifies which systems are close to the boundary and explains the regional structure of merger probability.
  • Because the merger boundary is irregular and likely fractal, any finite sample leaves unresolved fine structure, and reported merger fractions in narrow inclination bands (e.g., the trough near 60 degrees) should be treated as resolved only at the sampled resolution.

Reading between the lines

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

  • The paper's accuracy numbers are measured on a test set with roughly 70% nonmergers drawn from the MCMC-weighted sample; a uniform population-synthesis draw could shift precision and recall, so the network should be recalibrated on the target population before rate predictions.
  • If the mass-asymmetry trend is real, the hierarchical triple channel predicts a detectable excess of unequal-mass inner binaries in the gravitational-wave catalog compared to isolated-binary channels.
  • Because semisecular corrections can enhance eccentricity excitation, the true merger-conducive volume may be larger than this map shows, making the map a lower bound on the hierarchical triple channel's contribution.
  • The 54- and 64-degree inclination peaks, if they survive in a higher-order secular treatment, would be a dynamical signature (possibly a resonance or chaos boundary) that could be probed by targeted N-body runs.
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

4 major / 4 minor

Summary. The manuscript presents a large-scale numerical exploration of hierarchical triple black hole systems using a modified version of the JADE secular code that incorporates gravitational-wave energy loss. An adaptive MCMC scheme, biased toward the merger/nonmerger boundary via a nearest-neighbor scoring function, is used to simulate ~14.5 million unique configurations spanning a seven-dimensional parameter space of masses, semimajor axes, outer eccentricity, and mutual inclination. The authors identify qualitative trends for merger-conducive regions (asymmetric inner masses, moderate inner separations, small outer separations, high outer eccentricities, and higher inclinations) and classify nonmerging systems into four categories based on GW emission and ZLK activity. They additionally train an open-source neural network classifier on the JADE outputs, reporting 94.7% test accuracy, 99.0% AUC, and 99.7% accuracy for high-confidence predictions, and validate JADE against TSUNAMI N-body simulations for 1,000 select systems.

Significance. If the quantitative claims are robust, this would be the first systematic 7D map of the triple-BH merger boundary and a useful fast surrogate for population synthesis. The physical trends are consistent with established ZLK/octupole theory and are backed by perfect (300/300) N-body agreement in far-from-boundary regimes, giving confidence in the qualitative classification and in the decision to exclude the high-inclination and high-eccentricity regions where the secular approximation fails. The released code and dataset are reproducibility assets, and the parameter-space characterization of nonmerging channels is a novel contribution. However, as detailed below, the headline numerical accuracies currently overstate the reliability of the model relative to true triple dynamics, and the merger fractions presented in Figure 3 are affected by the deliberate sampling bias of the MCMC without any described reweighting.

major comments (4)
  1. [Sec. 4.2 and Fig. 3] The merger fractions plotted in Figure 3 and the percentages quoted in Section 4.2 (e.g., the categories of nonmerging systems) are computed from the adaptive MCMC sample described in Section 3.2, which is deliberately concentrated near the f≈0.5 boundary via the scoring function g(x) of Eq. (9). No reweighting to a uniform prior over the 7D box is described, and the retained duplicates further weight the sample toward high-scoring boundary regions. Consequently, the 'fraction of systems that merge' in a given cell reflects the proposal density rather than the physical merger probability under a uniform parameter distribution. The authors should either compute importance weights (inverse of the effective proposal density) or restrict the quantitative merger-fraction claims to the original 300,000-system grid, and otherwise explicitly label Fig. 3 as reporting boundary-focused sample fractions rather than uniform-prior probabilities.
  2. [Sec. 4.3 and Table 1] The neural network's reported test performance (94.7% accuracy, 99.0% AUC, 99.7% high-confidence accuracy) is measured on a held-out subset of the same JADE-generated labels used for training. Section 4.1 shows that JADE itself agrees with TSUNAMI N-body for only 79% of stable boundary-region systems (382/482) and 87% overall outside the excluded regimes. The network therefore measures its own agreement with the secular model, not with physical triple dynamics, and it cannot correct classification errors already present in its training labels. The abstract and Section 5 present the 99.7% figure without this essential caveat, which is misleading. Please either (a) rephrase all accuracy metrics as 'agreement with the JADE model,' (b) validate the neural network on the existing N-body validation sample and report the outcome, or (c) propagate the secular-model error into the stated NN accuracies so that the headline number reflects expected physical accuracy.
  3. [Sec. 4.1] The overall validation accuracy of 87% for stable systems outside the excluded regimes is an unweighted average over 300 far-from-boundary systems (100% agreement) and 482 boundary systems (79% agreement). Presenting this as a single number obscures the substantially lower reliability precisely in the merger/nonmerger boundary region that the MCMC is designed to map. The paper should report the boundary-subset accuracy (79%) separately whenever judging the robustness of the boundary map, rather than only in the detailed text, and should avoid giving the impression that the 87% figure characterizes the regions where the neural network and merger fractions are most consequential.
  4. [Sec. 4.2 and Sec. 5] Section 5 acknowledges the absence of semisecular corrections and the exclusion of the high-inclination and high-eccentricity regimes, yet the discussion in Section 4.2 states boundary-region conclusions without these qualifications, such as 'Systems with aouter ≳ 10 ainner almost universally merge.' Given that only 79% of boundary systems are correctly classified by JADE, the paper should either quantify the systematic uncertainty in such boundary-region statements (e.g., by applying the 79%/87% agreement rates to the reported fractions) or explicitly label them as predictions of the hexadecapolar secular approximation rather than as established physical facts. This caveat should also be reflected in the abstract, which currently presents the NN accuracy without indicating that it is a self-consistency measure.
minor comments (4)
  1. [Sec. 3.2] The parameter ranges are stated as ainner ∈ [1,200] AU and aouter ∈ [100,10,000] AU with the constraint aouter > 10 ainner, but the text does not specify how this constraint is imposed during sampling; since a large fraction of the naive box violates it (e.g., ainner = 200 AU, aouter = 500 AU), the effective parameter volume is much smaller than the nominal 7D box. Please clarify the sampling procedure and report the effective volume or the fraction of proposed points that satisfy the hierarchical condition.
  2. [Sec. 3.2] The sentence 'if the ZLK mechanism is active, it will oscillate through its full range regardless of the initial value' is too strong, since the maximum eccentricity in the octupole regime can depend on the initial eccentricity and on secular resonances. While setting einner = 0 is a reasonable and common choice for the purposes of this systematic sweep, the justification should be softened or referenced to the relevant literature (e.g., Naoz 2016; Liu et al. 2015).
  3. [Abstract and Sec. 4.1] The abstract states '87% qualitative agreement' with N-body simulations, but Section 4.1 reports a quantitative binary-outcome agreement of 87% excluding two excluded regions, and a 79% agreement for the boundary subset. The word 'qualitative' is unnecessarily vague for a percentage; please use '87% agreement in merger outcome' and, in the abstract, also mention the boundary-subset accuracy or at least the exclusion of the two problematic regimes.
  4. [Sec. 3.3] The description of input normalization says only that parameters were normalized to [0,1]; for reproducibility, please specify whether this is a min–max normalization over the full 7D box and, if so, over the raw ranges or the effective hierarchical subset.

Circularity Check

1 steps flagged · score 4.0 of 10

Neural-network accuracy is measured against the same JADE code that produced its labels, so the 95%/99.7% headline is a surrogate self-consistency score, not an independent physical prediction.

  1. fitted input called prediction [Sections 3.2-3.3 (data generation and NN training) and Section 4.3 (NN performance, Table 1)]
    "For each system, we computed whether it merged within 14 Gyr or not, assigning a value of 1 (merger) or 0 (no merger). ... we trained a neural network on the results of our MCMC exploration. After removing duplicates, we had 14,456,391 unique systems with known outcomes, which we split into training (60%), validation (20%), and test (20%) sets."

    The 'known outcomes' used for training and as the test-set 'Actual Outcome' (Section 4.3, Table 1) are exactly the M(x) merger flags produced by the modified JADE code in Section 3.2. The reported 94.7% accuracy, 99.0% AUC, and 99.7% high-confidence accuracy therefore quantify the network's fidelity to JADE, not to physical triple dynamics. The Abstract and Section 5 state this as 'predicts merger outcomes' without the qualifier 'as classified by JADE.' This matters because Section 4.1 shows JADE matches the TSUNAMI N-body code on only 79% of stable systems in the uncertain boundary region (382/482); the headline NN figures are a self-consistency score of the surrogate, so they do not independently certify physical predictive accuracy.

full rationale

The physical parameter-space map is not circular: it is produced by the modified JADE secular model, and the paper independently checks that model against the TSUNAMI N-body integrator on 1,000 systems (Section 4.1). That external check, despite the boundary-region 79% agreement, provides a genuine anchor for the trends in Section 4.2; the exclusions of high-inclination and high-outer-eccentricity regions are disclosed rather than hidden. The self-citation to Attia et al. (2021) for JADE's secularization is not load-bearing because the code is open source and the current paper supplies direct N-body validation. The only circular step is the neural-network evaluation: the 'actual outcome' in its confusion matrix is the same modified-JADE output on which the network was trained, so the quoted accuracy statistics are internal reproducibility scores rather than independent physical validation. I do not count the adaptive-MCMC oversampling without reweighting as circularity, but it is a separate sampling-bias concern for the quantitative merger fractions in Figure 3. Overall, the central derivation has independent content, so the circularity score is moderate, not severe.

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

The paper introduces no new free parameters or postulated entities. It relies on the standard secular Hamiltonian truncated at hexadecapolar order, Peters' GW formulas, the prograde/retrograde symmetry of ZLK, a simple hierarchy criterion, and the assumption that initial inner eccentricity is irrelevant. The main unproven domain assumptions are the sufficiency of the truncated secular model and the completeness of the 7-parameter description.

assumptions (6)
  • domain assumption The secular Hamiltonian truncated at hexadecapolar order (n=4) is sufficient for classifying merger outcomes.
    Section 3.1 relies on the JADE truncation following Beust et al. 2012; higher-order and semisecular terms are omitted.
  • standard math Peters (1964) formulas for GW emission describe the inner binary's angular momentum and eccentricity evolution.
    Equations (4) and (5) are taken as exact input from general relativity; the outer orbit's GW emission is neglected.
  • domain assumption ZLK dynamics are symmetric about 90 degrees mutual inclination for the parameter space, so retrograde orbits need not be simulated.
    Section 3.2 invokes Anderson et al. 2017 and Mangipudi et al. 2022 to exclude retrograde inclinations.
  • domain assumption The hierarchy criterion a_outer > 10 a_inner is sufficient for dynamical stability except for about 2% contamination.
    Section 4.1 acknowledges that more sophisticated stability criteria exist but accepts the simple criterion for computational efficiency.
  • domain assumption The initial inner eccentricity need not be varied because an active ZLK mechanism drives eccentricity through its full range regardless of the starting value.
    Section 3.2 states this assumption without a proof for all parameter combinations.
  • domain assumption Merger means the inner binary coalesces within 14 Gyr.
    A conventional choice in the field, used to define the binary merger label throughout the paper.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Exploring the parameter space of hierarchical triple black hole systems." pith.science (2026). https://pith.science/paper/BZUQOTMM

@misc{pith2026250622519,
  author       = {Pith},
  title        = {Pith review of: Exploring the parameter space of hierarchical triple black hole systems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BZUQOTMM}},
  note         = {Machine review of arXiv:2506.22519}
}
abstract

We present a comprehensive exploration of hierarchical triple black hole (BH) systems to address the "initial separation" problem in gravitational wave astrophysics. This problem arises because isolated BH binaries must have extremely small initial separations to merge within a Hubble time via gravitational wave (GW) emission alone, separations at which their stellar progenitors would have merged prematurely. Using a modified JADE secular code incorporating GW energy loss, we systematically investigate a seven-dimensional parameter space: masses of three BHs (5-100 $M_\odot$ inner binary, 1-200 $M_\odot$ tertiary), inner/outer semimajor axes (1-200 AU and 100-10,000 AU), outer orbit eccentricity (0-0.9), and mutual inclination (40{\deg}-80{\deg}). We employed an adaptive MCMC approach sampling the merger/nonmerger transition boundary across nearly 15 million simulations. Results reveal merger-conducive regions correspond to asymmetric inner binary masses, large inner separations where von Zeipel-Lidov-Kozai (ZLK) mechanism operates effectively without relativistic precession suppression, small outer separations providing stronger perturbations, and large outer eccentricities bringing the tertiary closer at pericenter. Merger probability correlates positively with mutual inclination. We developed a classification scheme for nonmerging systems based on GW emission and ZLK oscillations. A trained neural network predicts merger outcomes with 99% ROC score and 95% accuracy (99.7% for high-confidence predictions), enabling rapid population synthesis. Validation against N-body integrations showed 87% qualitative agreement, confirming our methodology captures essential triple BH dynamics while enabling unprecedented-scale exploration of configurations resolving the initial separation problem.

Figures

Figures reproduced from arXiv: 2506.22519 by the authors.

Figure 1
Figure 1. Generic orbital configuration of a hierarchical triple system, as investigated in this study. 3. Methods 3.1. Secular evolution with the JADE code We use a modified version of the JADE1 code (Attia et al. 2021), which was originally developed to simulate the secular evolu￾tion of hierarchical triple systems in the context of planetary dynamics. The code treats the three-body problem using a sec￾ular approximation, w… view at source ↗
Figure 2
Figure 2. Comparative evolution of inner semimajor axis (top) and inner eccentricity (bottom left), with Fourier transform of inner eccentricity (bottom right) including position of peak frequencies (dotted lines), between N-body (black) and secular (purple) codes for two similar initial configurations: M1 “ 68.7 Md, M2 “ 17.0 Md, M3 “ 92.0 Md, ainner “ 88.0 AU, aouter “ 6.82 kAU, eouter “ 0.670, imut “ 78.3 ˝ . secular code … view at source ↗
Figure 3
Figure 3. Fraction of systems that merge in each cell for each pair of parameters, or in each bin for each single parameter. The color ranges from purple (no systems merge) through white („ 50% merger rate) to yellow (all systems merge) [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Distribution of three categories of nonmerging systems in the parameter space. Each category is represented by a base color: cyan (GW, ZLK), magenta (no GW, ZLK), and yellow (GW, no ZLK). For each pair of parameters, cells are color-coded according to the combination o…
Figure 6
Figure 6. Figure 6: Histograms of true positives, true negatives, false positives, and false negatives as functions of prediction confidence cpyˆq. high-confidence systems, the model achieves a positive predic￾tive value of 99.7% and a negative predictive value of 99.6%. In practical term…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

70 extracted references · 48 canonical work pages

  1. [1]

    D., Acernese, F., et al

    Abbott, R., Abbott, T. D., Acernese, F., et al. 2023, Physical Review X, 13, 041039

    2023

  2. [2]

    R., Lai, D., & Storch, N

    Anderson, K. R., Lai, D., & Storch, N. I. 2017, MNRAS, 467, 3066

    work page 2017

  3. [3]

    L., et al

    Antonini, F., Chatterjee, S., Rodriguez, C. L., et al. 2016, ApJ, 816, 65

    2016

  4. [4]

    2017, MNRAS, 464, L36

    Askar, A., Szkudlarek, M., Gondek-Rosi´nska, D., Giersz, M., & Bulik, T. 2017, MNRAS, 464, L36

    2017

  5. [5]

    2021, A&A, 647, A40

    Attia, M., Bourrier, V ., Eggenberger, P., et al. 2021, A&A, 647, A40

    work page 2021

  6. [6]

    2010, MNRAS, 402, 371

    Banerjee, S., Baumgardt, H., & Kroupa, P. 2010, MNRAS, 402, 371

    work page 2010

  7. [7]

    2023, arXiv e-prints, arXiv:2302.10350

    Bartos, I., Rosswog, S., Gayathri, V ., et al. 2023, arXiv e-prints, arXiv:2302.10350

    arXiv 2023

  8. [8]

    2021, Phys

    Bern, Z., Luna, A., Roiban, R., Shen, C.-H., & Zeng, M. 2021, Phys. Rev. D, 104, 065014

    work page 2021

Show all 70 references
  1. [9]

    2012, A&A, 545, A88

    Beust, H., Bonfils, X., Montagnier, G., Delfosse, X., & Forveille, T. 2012, A&A, 545, A88

    2012

  2. [10]

    M., Stevance, H

    Briel, M. M., Stevance, H. F., & Eldridge, J. J. 2023, MNRAS, 520, 5724

    2023

  3. [11]

    C., & Michaely, E

    Britt, D., Johanson, B., Wood, L., Miller, M. C., & Michaely, E. 2021, MNRAS, 505, 3844

    2021

  4. [12]

    Chandramouli, R. S. & Yunes, N. 2022, Phys. Rev. D, 105, 064009 de Mink, S. E. & Mandel, I. 2016, MNRAS, 460, 3545

    2022

  5. [13]

    J., Dempsey, A

    Dittmann, A. J., Dempsey, A. M., & Li, H. 2024, ApJ, 964, 61 du Buisson, L., Marchant, P., Podsiadlowski, P., et al. 2020, MNRAS, 499, 5941 Duchêne, G. & Kraus, A. 2013, ARA&A, 51, 269

    2024

  6. [14]

    1916, Sitzungsberichte der Königlich Preussischen Akademie der Wissenschaften, 688

    Einstein, A. 1916, Sitzungsberichte der Königlich Preussischen Akademie der Wissenschaften, 688

    1916

  7. [15]

    1918, Sitzungsberichte der Königlich Preussischen Akademie der Wissenschaften, 154

    Einstein, A. 1918, Sitzungsberichte der Königlich Preussischen Akademie der Wissenschaften, 154

    1918

  8. [16]

    & Kocsis, B

    Fragione, G. & Kocsis, B. 2018, Phys. Rev. Lett., 121, 161103

    2018

  9. [17]

    Gallegos-Garcia, M., Berry, C. P. L., Marchant, P., & Kalogera, V . 2021, ApJ, 922, 110

    2021

  10. [18]

    & Perets, H

    Generozov, A. & Perets, H. B. 2024, ApJ, 964, 83 Gondán, L. 2023, MNRAS, 519, 1856

    2024

  11. [19]

    Hamers, A. S. 2021, MNRAS, 500, 3481

    2021

  12. [20]

    Hao, W., Kouwenhoven, M. B. N., Spurzem, R., et al. 2024, MNRAS, 527, 10705

    2024

  13. [21]

    Harrington, R. S. 1968, AJ, 73, 190

    1968

  14. [22]

    Hartl, M. D. & Buonanno, A. 2005, Phys. Rev. D, 71, 024027

    2005

  15. [23]

    A., & Suto, Y

    Hayashi, T., Trani, A. A., & Suto, Y . 2022, ApJ, 939, 81

    2022

  16. [24]

    A., & Suto, Y

    Hayashi, T., Trani, A. A., & Suto, Y . 2023, ApJ, 943, 58 3 https://github.com/maraattia/TripleBHMergerPredictor Article number, page 10 of 11 M. Attia & Y . Sibony: Hierarchical Triple Black Hole Parameter Space

    2023

  17. [25]

    A., & Dosopoulou, F

    Hoang, B.-M., Naoz, S., Kocsis, B., Rasio, F. A., & Dosopoulou, F. 2018, ApJ, 856, 140

    2018

  18. [26]

    2017, MNRAS, 468, 5020

    Inayoshi, K., Hirai, R., Kinugawa, T., & Hotokezaka, K. 2017, MNRAS, 468, 5020

    2017

  19. [27]

    2013, A&A Rev., 21, 59

    Ivanova, N., Justham, S., Chen, X., et al. 2013, A&A Rev., 21, 59

    2013

  20. [28]

    2011, Phys

    Katz, B., Dong, S., & Malhotra, R. 2011, Phys. Rev. Lett., 107, 181101

    2011

  21. [29]

    E., Will, C

    Kidder, L. E., Will, C. M., & Wiseman, A. G. 1993, Phys. Rev. D, 47, R4183

    1993

  22. [30]

    & Nakai, H

    Kinoshita, H. & Nakai, H. 1999, Celestial Mechanics and Dynamical Astronomy, 75, 125

    1999

  23. [31]

    R., Hut, P., & McMillan, S

    Kulkarni, S. R., Hut, P., & McMillan, S. 1993, Nature, 364, 421

    1993

  24. [32]

    & Trani, A

    Lalande, F. & Trani, A. A. 2022, ApJ, 938, 18

    2022

  25. [33]

    G., & Zhang, B

    Lepp, S., Martin, R. G., & Zhang, B. 2023, ApJ, 958, L23

    2023

  26. [34]

    Lidov, M. L. 1962, Planet. Space Sci., 9, 719

    1962

  27. [35]

    & Lai, D

    Liu, B. & Lai, D. 2020, Phys. Rev. D, 102, 023020

    2020

  28. [36]

    & Lai, D

    Liu, B. & Lai, D. 2021, MNRAS, 502, 2049

    2021

  29. [37]

    J., & Lai, D

    Liu, B., Muñoz, D. J., & Lai, D. 2015, MNRAS, 447, 747

    2015

  30. [38]

    Liu, S., Wang, L., Hu, Y .-M., Tanikawa, A., & Trani, A. A. 2024, MNRAS, 533, 2262

    2024

  31. [39]

    & Soker, N

    Livio, M. & Soker, N. 1988, ApJ, 329, 764

    1988

  32. [40]

    & Broekgaarden, F

    Mandel, I. & Broekgaarden, F. S. 2022, Living Reviews in Relativity, 25, 1

    2022

  33. [41]

    & de Mink, S

    Mandel, I. & de Mink, S. E. 2016, MNRAS, 458, 2634

    2016

  34. [42]

    & Farmer, A

    Mandel, I. & Farmer, A. 2022, Phys. Rep., 955, 1

    2022

  35. [43]

    A., & Mandel, I

    Mangipudi, A., Grishin, E., Trani, A. A., & Mandel, I. 2022, ApJ, 934, 44

    2022

  36. [44]

    H., Rantala, A., Naab, T., & Liao, S

    Mannerkoski, M., Johansson, P. H., Rantala, A., Naab, T., & Liao, S. 2021, ApJ, 912, L20

    2021

  37. [45]

    2018, arXiv e-prints, arXiv:1809.09130

    Mapelli, M. 2018, arXiv e-prints, arXiv:1809.09130

    2018 arXiv

  38. [46]

    2021, in Handbook of Gravitational Wave Astronomy, ed

    Mapelli, M. 2021, in Handbook of Gravitational Wave Astronomy, ed. C. Bambi, S. Katsanevas, & K. D. Kokkotas, 16

    2021

  39. [47]

    2021, Symmetry, 13, 1678

    Mapelli, M., Santoliquido, F., Bouffanais, Y ., et al. 2021, Symmetry, 13, 1678

    2021

  40. [48]

    M., & Moriya, T

    Marchant, P., Langer, N., Podsiadlowski, P., Tauris, T. M., & Moriya, T. J. 2016, A&A, 588, A50

    2016

  41. [49]

    Mardling, R. A. & Aarseth, S. J. 2001, MNRAS, 321, 398

    2001

  42. [50]

    Martinez, M. A. S., Fragione, G., Kremer, K., et al. 2020, ApJ, 903, 67

    2020

  43. [51]

    Miller, M. C. 2016, General Relativity and Gravitation, 48, 95

    2016

  44. [52]

    A., & Umbreit, S

    Morscher, M., Pattabiraman, B., Rodriguez, C., Rasio, F. A., & Umbreit, S. 2015, ApJ, 800, 9

    2015

  45. [53]

    Murray, C. D. & Dermott, S. F. 1999, Solar System Dynamics

    1999

  46. [54]

    & Katz, B

    Mushkin, J. & Katz, B. 2020, MNRAS, 498, 665

    2020

  47. [55]

    2013, ApJ, 773, 187

    Naoz, S., Kocsis, B., Loeb, A., & Yunes, N. 2013, ApJ, 773, 187

    2013

  48. [56]

    1976, in IAU Symposium, V ol

    Paczynski, B. 1976, in IAU Symposium, V ol. 73, Structure and Evolution of Close Binary Systems, ed. P. Eggleton, S. Mitton, & J. Whelan, 75

    1976

  49. [57]

    Peters, P. C. 1964, Physical Review, 136, 1224 Portegies Zwart, S. F. & McMillan, S. L. W. 2000, ApJ, 528, L17

    1964

  50. [58]

    Postnov, K. A. & Yungelson, L. R. 2014, Living Reviews in Relativity, 17, 3

    2014

  51. [59]

    2021, MNRAS, 505, 663

    Riley, J., Mandel, I., Marchant, P., et al. 2021, MNRAS, 505, 663

    2021

  52. [60]

    L., Haster, C.-J., Chatterjee, S., Kalogera, V ., & Rasio, F

    Rodriguez, C. L., Haster, C.-J., Chatterjee, S., Kalogera, V ., & Rasio, F. A. 2016, ApJ, 824, L8

    2016

  53. [61]

    & Hernquist, L

    Sigurdsson, S. & Hernquist, L. 1993, Nature, 364, 423

    1993

  54. [62]

    & Tremaine, S

    Silsbee, K. & Tremaine, S. 2017, ApJ, 836, 39

    2017

  55. [63]

    & Klencki, J

    Stegmann, J. & Klencki, J. 2025, arXiv e-prints, arXiv:2506.09121

    2025

  56. [64]

    2021, MNRAS, 505, 3681

    Su, Y ., Liu, B., & Lai, D. 2021, MNRAS, 505, 3681

    2021

  57. [65]

    2021, Universe, 7, 352

    Tokovinin, A. 2021, Universe, 7, 352

    2021

  58. [66]

    A., Leigh, N

    Trani, A. A., Leigh, N. W. C., Boekholt, T. C. N., & Portegies Zwart, S. 2024, A&A, 689, A24

    2024

  59. [67]

    A., Rastello, S., Di Carlo, U

    Trani, A. A., Rastello, S., Di Carlo, U. N., et al. 2022, MNRAS, 511, 1362

    2022

  60. [68]

    Trani, A. A. & Spera, M. 2023, in The Predictive Power of Computational Astro- physics as a Discover Tool, ed. D. Bisikalo, D. Wiebe, & C. Boily, V ol. 362, 404–409

    2023

  61. [69]

    2023, MNRAS, 522, 937 van Son, L

    Tremaine, S. 2023, MNRAS, 522, 937 van Son, L. A. C., de Mink, S. E., Renzo, M., et al. 2022, ApJ, 940, 184 von Zeipel, H. 1910, Astronomische Nachrichten, 183, 345

    2023

  62. [70]

    S., Mardling, R

    Vynatheya, P., Hamers, A. S., Mardling, R. A., & Bellinger, E. P. 2022, MNRAS, 516, 4146 Article number, page 11 of 11

    2022

Pith tools

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