Pith. sign in

REVIEW 4 major objections 5 minor 74 references

Low-latency Forecasts of Kilonova Light Curves for Rubin and ZTF

T0 review · 4 major / 5 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 The paper shows that a bidirectional LSTM trained on simulated gravitational-wave alerts can forecast kilonova light curves in ZTF and Rubin filters, with a test MSE of 0.19 (ZTF) and 0.22 (Rubin), and that adding ejecta mass as an input…

desk verdict A reproducible emulator of one restricted family of simulated kilonova light curves, with honest reporting and a headline claim that runs ahead of the demonstrated domain. read the letter →

arxiv 2507.11785 v1 pith:J3YHRDYD submitted 2025-07-15 astro-ph.HE

Natalya Plestkova , Niharika Sravan , R. Weizmann Kiendrebeogo , Michael W. Coughlin , Derek Davis , Andrew Toivonen , Theophile Jegou du Laz , Tomás Ahumada
show 6 more authors
Tyler Barna George Helou Roger Smith Ben Rusholme Russ R. Laher Ashish A. Mahabal
This is my paper · ORCID
classification astro-ph.HE
keywords kilonovagravitationalwaveslightcurveforecastingLSTMneuralnetworksZTFRubinObservatorymulti-messengerastronomy
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 presents a public machine-learning tool that forecasts kilonova light curves from the low-latency alert data that gravitational-wave observatories issue minutes after a merger candidate is found. The authors train a bidirectional LSTM on simulated binary neutron star and neutron star-black hole mergers, using only alert-derived features such as distance, sky-localization area, and the probabilities HasNS, HasRemnant, HasMassGap, and PAstro, and show it predicts simulated light curves across ZTF g/r/i and Rubin u/g/r/i/y/z bands with test MSE 0.19 and 0.22 respectively. They argue this is accurate enough to help plan electromagnetic follow-up by estimating whether a kilonova will be visible, how bright it will peak, and how fast it will fade. They also find that adding ejecta mass as an input lowers the MSE to about 0.1, while including full skymap images does not help. A sympathetic reader would take the core claim to be that low-latency GW alert parameters contain enough information to support useful kilonova brightness forecasts before any detailed inference is available.

What carries the argument

The mechanism is a bidirectional LSTM that maps a vector of six low-latency alert features — distance, 90% sky-localization area, HasNS, HasRemnant, HasMassGap, and PAstro — to a 90-unit output representing 30 time steps across three ZTF filters, or a 180-unit output for six Rubin filters. The training labels come from the NMMA framework with POSSIS radiative-transfer models, which convert binary masses, tidal deformability, and spins into ejecta masses and multiband light curves for each simulated merger. A hybrid CNN-LSTM variant additionally encodes Bayestar skymaps as 500x1000 three-channel images, while an augmented model adds ejecta mass as a scalar feature, and the feature-importance analysis shows that this single physically motivated input carries most of the predictive gain.

What would settle it

Feed the low-latency alert parameters of GW170817 (or a future well-localized BNS with a detected kilonova) into the public model and compare its forecast to the observed AT2017gfo light curves in the same filters; a large mismatch would show that the model tracks the NMMA/POSSIS training distribution rather than real kilonova behavior.

Watch

Extended reading notes

Core claim

The central claim is that a bidirectional long-short-term memory network, trained exclusively on simulated IGWN alert features, can serve as a low-latency emulator of kilonova light curves produced by the NMMA/POSSIS framework, and that this emulator is accurate enough to inform follow-up decisions. The model achieves a test mean squared error of 0.19 across ZTF g, r, and i filters and 0.22 across six Rubin filters, with more than half of the test light curves having individual MSE below 0.2. The authors verify the model against candidates followed up during O4a and O4b, finding reasonable agreement for S230627c and no match for 250206dm. They further show that a hybrid CNN-LSTM model using full skymap images does not improve over the LSTM-only model, whereas adding the physically motivated ejecta-mass feature improves the MSE to 0.1 on ZTF and 0.10 on Rubin filters, identifying a clear path for future alert products.

Load-bearing premise

The training labels are simulated light curves from the NMMA/POSSIS framework, so the reported accuracy measures agreement with those simulations; if the simulations misrepresent real kilonova brightness or evolution, the forecast error on real events will be larger than the reported MSE.

Editorial extensions

If this is right

  • Observers can run the public model on an incoming IGWN alert and receive a first estimate of kilonova brightness and evolution across ZTF or Rubin filters within seconds of the alert.
  • Follow-up planning can use the forecasts to decide whether an event is worth chasing and which filters and exposure times to use, including for high-FAR events that current selection cuts discard.
  • For Rubin/LSST, the model provides a preview in six bands, including near-infrared y and z where the simulated KNe are best predicted, which can inform Target of Opportunity program design.
  • If the LVK begins releasing coarse ejecta-mass estimates in public alerts, the model's expected error should drop to roughly half of its current level, based on the demonstrated improvement to MSE 0.1.
  • The skymap CNN experiment suggests that full sky-localization images add little beyond the scalar area(90), so simpler alert features suffice for light-curve forecasting.

Reading between the lines

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

  • Because the MSE is measured against NMMA/POSSIS simulations, the reported numbers bound how well the model reproduces those simulations, not how well it predicts a real kilonova; the tool's value on the sky depends on how faithful the simulations are to real events.
  • The same architecture could be retrained on other synthetic light-curve libraries or on real kilonovae as they accumulate, and the feature-importance result suggests that any future alert product carrying an ejecta-mass estimate would yield the largest immediate accuracy gain.
  • A natural extension is to use the forecast not as a point prediction but as a prior for a matching stage that cross-correlates forecast light curves against alert-stream candidates, an idea the authors gesture toward with their planned integration into platforms like SkyPortal.
  • The model's tendency to under-predict brightness for faint, unusual light curves, if it persists on real data, would bias follow-up toward deeper searches, which is the safer direction for discovery.
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 / 5 minor

Summary. The paper presents a public machine-learning tool that forecasts kilonova light curves in ZTF and Rubin/LSST filters from low-latency IGWN alert features (distance, area(90), HasNS, HasRemnant, HasMassGap, PAstro). A bidirectional LSTM is trained on NMMA/POSSIS simulated light curves from 7,985 O4 BNS/NSBH events for ZTF and 4,223 O5 events for Rubin, reporting test MSE of 0.19 (ZTF, R2=0.82) and 0.22 (Rubin, R2=0.68). The authors also test a CNN-LSTM variant using skymaps, find no improvement, and show that adding ejecta mass as a feature improves MSE to about 0.10-0.11. They compare forecasts to O4a/O4b follow-up candidates, with mixed qualitative agreement. The paper emphasizes the operational goal of helping plan electromagnetic follow-up of gravitational-wave events.

Significance. If the central claim is properly scoped, the tool is a potentially useful, fast surrogate for NMMA/POSSIS light-curve generation from low-latency GW alert parameters, with clear practical value for target-of-opportunity follow-up planning. The paper's strengths include a public GitHub repository with pretrained models, reproducible training scripts, and use of a held-out simulated test set, which makes the internal MSE numbers checkable. The main scientific value is the demonstration that low-latency alert features carry enough information to reproduce the conditional average simulated kilonova light curve, and the analysis of which added features help. However, the applicability to real kilonovae is not established by the present evidence: the training population is deliberately restricted in ejecta configuration, and the O4 validation includes no confirmed kilonova. The reported accuracy is therefore best interpreted as a simulation-to-simulation emulation accuracy for a restricted model family, not as a measured forecast skill for real events.

major comments (4)
  1. [Section 2, Training Data] The training set is generated with alpha = 0, zeta = 0.3, and epsilon = 0, turning off fallback and wind ejecta and allowing only 30% of the disk mass to be ejected. Since the six input features (distance, area(90), HasNS, HasRemnant, HasMassGap, PAstro) do not encode ejecta mass, velocity, composition, or viewing angle, the model can only learn the conditional average light curve of this restricted simulated population. The headline test MSE of 0.19 (ZTF) and 0.22 (Rubin) therefore measures agreement with a restricted simulation family, not expected error for a real kilonova with, for example, significant wind or fallback ejecta. The central claim in the abstract that the tool 'can help to add important information to help plan follow-up' needs to be scoped accordingly, or the training set needs to cover a broader ejecta parameter space.
  2. [Section 2, Eq. (1) and Eqs. (3)-(5)] The SNR-to-IFAR mapping log10(IFAR) = 2.357*SNR - 20.198 is a fitted relation from GWTC-3 BNS injections, and it is used together with assumed values of RCBC, the IFAR threshold, and the number of pipelines to synthesize PAstro. No uncertainty in the fitted slope and intercept is propagated, and the relation is applied to NSBH events and to O4/O5 sensitivity without validation. Because PAstro is one of the six input features, a biased mapping will directly bias the forecasted light curves. At minimum, the authors should compare their synthesized PAstro values against actual low-latency alert values for O4 events and quantify the sensitivity of the MSE to the assumed mapping parameters.
  3. [Section 4.4, Comparison to kilonova candidates] The O4 validation does not include a confirmed kilonova. For S230627c the text states the candidate may be a BBH merger with very small ejecta mass, and for 250206dm none of the candidates match the forecasts. The abstract's phrase 'We verify the performance of the model against merger events followed-up by ZTF' overstates what Section 4.4 can show; the section demonstrates only a qualitative comparison with candidate counterparts, not a validation of forecasting skill for real kilonovae. The manuscript should explicitly state this limitation and distinguish simulation-to-simulation accuracy from real-event transfer.
  4. [Section 4.3, CNN-LSTM and ejecta-mass feature] The improvement from adding ejecta mass (MSE 0.10-0.11) is close to tautological in this setup: the NMMA/POSSIS light curves are deterministic functions of the same ejecta properties that the added feature summarizes. The comparison is therefore not a fair test of whether an independent physical feature improves generalization. Moreover, ejecta mass is not currently available in low-latency GW alerts, as the authors acknowledge in the Discussion. The result should be framed as an upper bound on the possible gain from a future low-latency ejecta-mass product, not as a recommended input for the current tool.
minor comments (5)
  1. [Section 4.4 title] The section title contains a typo: 'Comparision' should be 'Comparison'.
  2. [Section 3, Model] The text says 'We use MC dropout (with value 0.1) to estimate mean and uncertainity of forecasted light curves'; 'uncertainity' should be 'uncertainty', and the description would benefit from stating how many stochastic forward passes were used for the uncertainty estimates.
  3. [Author affiliation line] The first author name appears corrupted as 'Nataly aPletskov a'; this should be corrected in the final manuscript.
  4. [Table 3] The per-filter rows for the Rubin model list g, r, i, u, y, z, which is not in wavelength order; reordering to u, g, r, i, z, y would make the trend in MSE easier to read.
  5. [Figure 12] The figure caption mentions a black line for FAR = 22, but the main text does not explain the origin or rationale of this specific FAR threshold; a sentence of context would help.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the held-out NMMA/POSSIS emulator test is self-contained; the ejecta-mass feature experiment is an expected feature-importance check, not a low-latency prediction.

full rationale

The central claim is that a bidirectional LSTM can emulate NMMA/POSSIS kilonova light curves from low-latency IGWN alert features. The reported MSEs are computed on a held-out test set drawn from the same simulation pipeline used for training, which is a legitimate surrogate-model evaluation: the POSSIS radiative-transfer code is an external simulator, and the LSTM parameters are not fit to the test light curves. None of the abstract's performance numbers is constructed from the model's own outputs, and no equation in the paper reduces to another by construction. The only mildly self-referential elements are (1) the use of the authors' own simulated IGWN alert products (Weizmann Kiendrebeogo et al. 2023) and the ejecta-parameter setup (alpha=0, zeta=0.3, epsilon=0) adopted from Toivonen et al. (2024), a self-citation that is a modeling choice for comparability rather than a load-bearing proof; and (2) the Section 4.3 ejecta-mass feature experiment, where the target light curves are causally generated from ejecta properties, so adding ejecta mass as an input partly inverts the simulator. The paper explicitly notes that ejecta masses are not available in low-latency, so that result is a feature-importance check, not a deployable prediction. Real-event validation against O4 candidates is acknowledged to be inconclusive (S230627c may be a BBH; no 250206dm candidates match), which is a correctness/validity caveat, not circularity. Overall circularity is minimal.

Assumptions & free parameters 9 free parameters · 5 assumptions · 0 invented entities

The model's free parameters are dominated by the empirical SNR-IFAR mapping and the restricted ejecta setup, both of which shape the synthetic alert features and light curves. The core axioms are the fidelity of the NMMA/POSSIS simulation suite and the representativeness of the reconstructed alert features. No new physical entities are introduced.

free parameters (9)
  • SNR-IFAR slope = 2.357
    Equation 1 fits log10(IFAR) = 2.357 * SNR - 20.198 to GWTC-3 injection medians; used to synthesize PAstro for all simulated training events.
  • SNR-IFAR intercept = -20.198
    Equation 1 intercept, same fit; no uncertainty quoted.
  • CBC rate RCBC = 35 per year per Gpc^3
    Assumed from O3 (Abbott et al. 2023) to compute PAstro; O4 may have a higher rate.
  • assumed IFAR threshold = 1 year
    Used in Eq. 4 to compute Rastro; not derived.
  • number of pipelines = 4
    Rnoise = 4 / IFAR assumes 4 independent discovery pipelines.
  • ejecta setup alpha = 0
    Foucart disk-wind parameter set to 0, turning off wind ejecta; this restricts the training set.
  • ejecta setup zeta = 0.3
    Disk mass ejection fraction fixed to 30%.
  • ejecta setup epsilon = 0
    Fallback ejecta parameter set to 0.
  • LSTM hyperparameters = units 400/300/200/150, LR 0.0003, batch 64, epochs 300
    Chosen by grid search on the validation set; the reported MSE depends on these choices.
assumptions (5)
  • domain assumption NMMA/POSSIS kilonova light curve models are an accurate representation of real kilonova emission.
    The training labels are generated by NMMA with POSSIS (Section 2). If these models are biased, the forecaster inherits the bias.
  • domain assumption The reconstructed PAstro distribution from the SNR-IFAR mapping matches the real IGWN alert PAstro distribution.
    PAstro for training is computed via Eq. 1-5 rather than taken from the actual low-latency pipeline; real alerts are used only for two events in Section 4.4.
  • domain assumption The injected population (uniform in comoving volume, SNR > 8) covers the parameter space of real detectable mergers.
    The training set is based on the IGWN User Guide simulations (Section 2); events outside this distribution (e.g., very bright nearby, or with strong wind ejecta) are poorly predicted (Section 4.1, Figure 6).
  • domain assumption MC dropout uncertainties are well calibrated.
    The 3-sigma bands in Figures 3, 7, 9 are from MC dropout; no calibration test is provided.
  • domain assumption Bayestar sky localization area(90) is an informative summary statistic for light curve forecasting.
    area(90) is used as a proxy for localization, but adding full skymaps did not improve performance (Section 4.3), suggesting the summary may be sufficient or the CNN processing is suboptimal.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Low-latency Forecasts of Kilonova Light Curves for Rubin and ZTF." pith.science (2026). https://pith.science/paper/J3YHRDYD

@misc{pith2026250711785,
  author       = {Pith},
  title        = {Pith review of: Low-latency Forecasts of Kilonova Light Curves for Rubin and ZTF},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/J3YHRDYD}},
  note         = {Machine review of arXiv:2507.11785}
}
read the original abstract

Follow-up of gravitational-wave events by wide-field surveys is a crucial tool for the discovery of electromagnetic counterparts to gravitational wave sources, such as kilonovae. Machine learning tools can play an important role in aiding search efforts. We have developed a public tool to predict kilonova light curves using simulated low-latency alert data from the International Gravitational Wave Network during observing runs 4 (O4) and 5 (O5). It uses a bidirectional long-short-term memory (LSTM) model to forecast kilonova light curves from binary neutron star and neutron star-black hole mergers in the Zwicky Transient Facility (ZTF) and Rubin Observatory's Legacy Survey of Space and Time filters. The model achieves a test mean squared error (MSE) of 0.19 for ZTF filters and 0.22 for Rubin filters, calculated by averaging the squared error over all time steps, filters, and light curves in the test set. We verify the performance of the model against merger events followed-up by the ZTF partnership during O4a and O4b. We also analyze the effect of incorporating skymaps and constraints on physical features such as ejecta mass through a hybrid convolutional neural network and LSTM model. Using ejecta mass, the performance of the model improves to an MSE of 0.1. However, using full skymap information results in slightly lower model performance. Our models are publicly available and can help to add important information to help plan follow-up of candidate events discovered by current and next-generation public surveys.

Figures

Figures reproduced from arXiv: 2507.11785 by the authors.

Figure 1
Figure 1. Mapping the signal-to-noise ratio to inverse false alarm rate using a large population of binary neutron stars injections. parison with the ejecta mass predictions from Toivonen et al. (2024) work and to ensure consistency in the training set for our ML models. To generate sky maps for simulated BNS and NSBH merg￾ers, we use Bayestar (Singer & Price 2016), extracting key parameters such as the sky localization area … view at source ↗
Figure 2
Figure 2. Architecture of the bidirectional LSTM model used for light curve predictions. of their ability to process sequential input in both forward and backward directions, allowing the model to identify pat￾terns and relationships between both future and previous ob￾servations (Schuster & Paliwal 1997). Dropout (Srivastava et al. 2014) is used for regularization, with a rate of 0.1 after each bidirectional LSTM layer. Foll… view at source ↗
Figure 3
Figure 3. Predicted (solid lines) and ground truth (dashed lines) light curves in ZTF filters using our trained model for a random KN in our test set. Shaded regions show 3-σ uncertainties [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (9 more)
Figure 4
Figure 4. Figure 4: Left: The best-predicted light curve with an MSE of 0.0027. The ground truth (dashed lines) and predicted light curves (solid lines) match closely. Right: The worst predicted light curve, with an MSE of 1.76 [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: MSE as a function of KN phase for ZTF filters. Small MSE values indicate better model performance. Early time fore￾casts (∼5 hours since merger) are unreliable. Forecasts in the r-filter are the best at early phases (≲ 2 days). which significantly improves its performa…
Figure 6
Figure 6. Figure 6: 2D histograms showing the peak magnitudes of KN light curves in the training dataset versus the MSE of the forecasted light curves in ZTF g (left), r (middle), and i (right) filters. The colorbar shows the number of training examples in each bin; larger counts are indi…
Figure 7
Figure 7. Figure 7: Predicted (solid lines) and ground truth (dashed lines) light curves in Rubin filters. Shaded regions show 3-σ uncertainties. and decline phases. This is particularly notable given that the most important feature, ejecta mass, is very small, since this candidate may be…
Figure 8
Figure 8. Figure 8: Similar to [PITH_FULL_IMAGE:figures/full_fig_p009_8.png]
Figure 9
Figure 9. Figure 9: Predicted mean (solid lines) and 3-σ uncertainities (shaded regions) of KN light curves associated with S240627c using low latency alert data [PITH_FULL_IMAGE:figures/full_fig_p010_9.png]
Figure 10
Figure 10. Figure 10: Same as [PITH_FULL_IMAGE:figures/full_fig_p010_10.png]
Figure 11
Figure 11. Figure 11: Feature importance of the fiducial model without ejecta mass (left) and with ejecta mass included (right) D. PEAK MAGNITUDE VS FAR [PITH_FULL_IMAGE:figures/full_fig_p015_11.png]
Figure 12
Figure 12. Figure 12: Peak magnitude vs FAR for O4 CBC detections. The purple line indicates the ZTF magnitude limit and the black line delineates FAR = 22. KNe associated with CBCs in the gray region are currently missed by follow-up programs using this FAR cut [PITH_FULL_IMAGE:figures/f…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

74 extracted references · 12 canonical work pages

  1. [1]

    P., Abbott, R., Abbott, T

    Abbott, B. P., Abbott, R., Abbott, T. D., et al. 2017a, Nature, 551, 85, doi: 10.1038/nature24471 —. 2017c, ApJL, 848, L13, doi: 10.3847/2041-8213/aa920c —. 2020a, ApJL, 892, L3, doi: 10.3847/2041-8213/ab75f5

    doi:10.1038/nature24471

  2. [2]

    D., Abraham, S., et al

    Abbott, R., Abbott, T. D., Abraham, S., et al. 2021, ApJL, 915, L5, doi: 10.3847/2041-8213/ac082e

    doi:10.3847/2041-8213/ac082e 2021

  3. [3]

    D., Acernese, F., et al

    Abbott, R., Abbott, T. D., Acernese, F., et al. 2023, Physical Review X, 13, 041039, doi: 10.1103/PhysRevX.13.041039 Abbott et al. 2017b, Phys. Rev. Lett., 119, 161101, doi: 10.1103/PhysRevLett.119.161101

    doi:10.1103/physrevx.13.041039 2023

  4. [4]

    P., Anand, S., et al

    Ahumada, T., Singer, L. P., Anand, S., et al. 2021, Nature Astronomy, 5, 917, doi: 10.1038/s41550-021-01428-7

    doi:10.1038/s41550-021-01428-7 2021

  5. [5]

    W., et al

    Ahumada, T., Anand, S., Coughlin, M. W., et al. 2024, PASP, 136, 114201, doi: 10.1088/1538-3873/ad8265

    doi:10.1088/1538-3873/ad8265 2024

  6. [6]

    2025, GRB Coordinates Network, 39228, 1

    Ahumada, T., Karambelkar, V ., Bellm, E., et al. 2025, GRB Coordinates Network, 39228, 1

    work page 2025

  7. [7]

    W., Kasliwal, M

    Anand, S., Coughlin, M. W., Kasliwal, M. M., et al. 2021, Nature Astronomy, 5, 46, doi: 10.1038/s41550-020-1183-3

    doi:10.1038/s41550-020-1183-3 2021

  8. [8]

    2025, PASP, 137, 034102, doi: 10.1088/1538-3873/adbfbc

    Andrade, C., Alserkal, R., Salazar Manzano, L., et al. 2025, PASP, 137, 034102, doi: 10.1088/1538-3873/adbfbc

    doi:10.1088/1538-3873/adbfbc 2025

Show all 74 references
  1. [9]

    2024, arXiv e-prints, arXiv:2411.04793, doi: 10.48550/arXiv.2411.04793

    Andreoni, I., Margutti, R., Banovetz, J., et al. 2024, arXiv e-prints, arXiv:2411.04793, doi: 10.48550/arXiv.2411.04793

    2024 doi arXiv

  2. [10]

    2020a, MNRAS, 492, 3904, doi: 10.1093/mnras/stz3142

    Antier, S., Agayeva, S., Aivazyan, V ., et al. 2020a, MNRAS, 492, 3904, doi: 10.1093/mnras/stz3142

    doi

  3. [11]

    2020b, MNRAS, 497, 5518, doi: 10.1093/mnras/staa1846

    Antier, S., Agayeva, S., Almualla, M., et al. 2020b, MNRAS, 497, 5518, doi: 10.1093/mnras/staa1846

    doi

  4. [12]

    A., et al

    Arcavi, I., Hosseinzadeh, G., Howell, D. A., et al. 2017, Nature, 551, 64, doi: 10.1038/nature24291

    2017 doi

  5. [13]

    C., Kulkarni, S

    Bellm, E. C., Kulkarni, S. R., Graham, M. J., et al. 2019, PASP, 131, 018002, doi: 10.1088/1538-3873/aaecbe

    2019 doi

  6. [14]

    2024, Classical and Quantum Gravity, 41, 085012, doi: 10.1088/1361-6382/ad3279

    Berbel, M., Miravet-Ten´es, M., Sharma Chaudhary, S., et al. 2024, Classical and Quantum Gravity, 41, 085012, doi: 10.1088/1361-6382/ad3279

    2024 doi

  7. [15]

    Biswas, B., Ishida, E. E. O., Peloton, J., et al. 2023, A&A, 677, A77, doi: 10.1051/0004-6361/202245340

    2023 doi

  8. [16]

    2019, MNRAS, 489, 5037, doi: 10.1093/mnras/stz2495

    Bulla, M. 2019, MNRAS, 489, 5037, doi: 10.1093/mnras/stz2495

    2019 doi

  9. [17]

    R., et al

    Chatterjee, D., Ghosh, S., Brady, P. R., et al. 2020, The Astrophysical Journal, 896, 54, doi: 10.3847/1538-4357/ab8e66

    2020 doi

  10. [18]

    2022, MNRAS, 509, 914, doi: 10.1093/mnras/stab3023

    Muthukrishna, D. 2022, MNRAS, 509, 914, doi: 10.1093/mnras/stab3023

    2022 doi

  11. [19]

    2022, MNRAS, 513, 5780, doi: 10.1093/mnras/stac1283

    Chattopadhyay, D., Stevenson, S., Broekgaarden, F., Antonini, F., & Belczynski, K. 2022, MNRAS, 513, 5780, doi: 10.1093/mnras/stac1283

    2022 doi

  12. [20]

    S., Toivonen, A., Waratkar, G., et al

    Chaudhary, S. S., Toivonen, A., Waratkar, G., et al. 2024, Proceedings of the National Academy of Science, 121, e2316474121, doi: 10.1073/pnas.2316474121

    2024 doi

  13. [21]

    2017, ApJL, 848, L19, doi: 10.3847/2041-8213/aa905c

    Chornock, R., Berger, E., Kasen, D., et al. 2017, ApJL, 848, L19, doi: 10.3847/2041-8213/aa905c

    2017 doi

  14. [22]

    W., Ahumada, T., Anand, S., et al

    Coughlin, M. W., Ahumada, T., Anand, S., et al. 2019, ApJL, 885, L19, doi: 10.3847/2041-8213/ab4ad8

    2019 doi

  15. [23]

    W., Bloom, J

    Coughlin, M. W., Bloom, J. S., Nir, G., et al. 2023, ApJS, 267, 31, doi: 10.3847/1538-4365/acdee1 12

    2023 doi

  16. [24]

    A., Foley, R

    Coulter, D. A., Foley, R. J., Kilpatrick, C. D., et al. 2017, Science, 358, 1556, doi: 10.1126/science.aap9811

    2017 doi

  17. [25]

    S., Villar, V

    Cowperthwaite, P. S., Villar, V . A., Scolnic, D. M., & Berger, E. 2019, ApJ, 874, 88, doi: 10.3847/1538-4357/ab07b6

    2019 doi

  18. [26]

    S., Berger, E., Villar, V

    Cowperthwaite, P. S., Berger, E., Villar, V . A., et al. 2017, ApJL, 848, L17, doi: 10.3847/2041-8213/aa8fc7

    2017 doi

  19. [27]

    M., Riddle, R., et al

    Dekany, R., Smith, R. M., Riddle, R., et al. 2020, PASP, 132, 038001, doi: 10.1088/1538-3873/ab4ca2

    2020 doi

  20. [28]

    2024, arXiv e-prints, arXiv:2408.06947, doi: 10.48550/arXiv.2408.06947 D´ıaz, M

    Desai, M., Chatterjee, D., Jhawar, S., et al. 2024, arXiv e-prints, arXiv:2408.06947, doi: 10.48550/arXiv.2408.06947 D´ıaz, M. C., Macri, L. M., Garcia Lambas, D., et al. 2017, ApJL, 848, L29, doi: 10.3847/2041-8213/aa9060

    2024 doi arXiv

  21. [29]

    W., Pang, P

    Dietrich, T., Coughlin, M. W., Pang, P. T. H., et al. 2020a, arXiv e-prints, arXiv:2002.11355. https: //arxiv.org/abs/2002.11355 —. 2020b, Science, 370, 1450, doi: 10.1126/science.abb4317

    2002 arXiv

  22. [30]

    T., Pahre, M

    Djorgovski, S., Soifer, B. T., Pahre, M. A., et al. 1995, ApJL, 438, L13, doi: 10.1086/187703

    1995 doi

  23. [31]

    R., Piro, A

    Drout, M. R., Piro, A. L., Shappee, B. J., et al. 2017, Science, 358, 1570, doi: 10.1126/science.aaq0049

    2017 doi

  24. [32]

    H., Vukobratovich, D., Andrew, J

    Elias, J. H., Vukobratovich, D., Andrew, J. R., et al. 1998, in Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, V ol. 3354, Infrared Astronomical Instrumentation, ed. A. M. Fowler, 555–565, doi: 10.1117/12.317281

    1998 doi

  25. [33]

    K., et al

    Fong, W., Berger, E., Blanchard, P. K., et al. 2017, ApJL, 848, L23, doi: 10.3847/2041-8213/aa9018

    2017 doi

  26. [34]

    2018, PhRvD, 98, 081501, doi: 10.1103/PhysRevD.98.081501

    Foucart, F., Hinderer, T., & Nissanke, S. 2018, PhRvD, 98, 081501, doi: 10.1103/PhysRevD.98.081501

    2018 doi

  27. [35]

    D., Smartt, S

    Fulton, M. D., Smartt, S. J., Huber, M. E., et al. 2025, arXiv e-prints, arXiv:2506.07082, doi: 10.48550/arXiv.2506.07082

    2025 doi arXiv

  28. [36]

    R., & Levan, A

    Gall, C., Hjorth, J., Rosswog, S., Tanvir, N. R., & Levan, A. J. 2017, ApJL, 849, L19, doi: 10.3847/2041-8213/aa93f9

    2017 doi

  29. [37]

    2017, ApJL, 848, L14, doi: 10.3847/2041-8213/aa8f41

    Goldstein, A., Veres, P., Burns, E., et al. 2017, ApJL, 848, L14, doi: 10.3847/2041-8213/aa8f41

    2017 doi

  30. [38]

    P., Cutter, R., Steeghs, D., et al

    Gompertz, B. P., Cutter, R., Steeghs, D., et al. 2020, MNRAS, 497, 726, doi: 10.1093/mnras/staa1845

    2020 doi

  31. [39]

    J., Kulkarni, S

    Graham, M. J., Kulkarni, S. R., Bellm, E. C., et al. 2019, PASP, 131, 078001, doi: 10.1088/1538-3873/ab006c GRA VITY Collaboration, Sanchez-Bermudez, J., Caratti O

    2019 doi

  32. [40]

    2021, A&A, 654, A97, doi: 10.1051/0004-6361/202039600

    Garatti, A., et al. 2021, A&A, 654, A97, doi: 10.1051/0004-6361/202039600

    2021 doi

  33. [41]

    S., Gomez, S., et al

    Hosseinzadeh, G., Cowperthwaite, P. S., Gomez, S., et al. 2019, ApJL, 880, L4, doi: 10.3847/2041-8213/ab271c

    2019 doi

  34. [42]

    2017, Science Bulletin, 62, 1433, doi: 10.1016/j.scib.2017.10.006 Ivezi´c, ˇZ., Kahn, S

    Hu, L., Wu, X., Andreoni, I., et al. 2017, Science Bulletin, 62, 1433, doi: 10.1016/j.scib.2017.10.006 Ivezi´c, ˇZ., Kahn, S. M., Tyson, J. A., et al. 2019, ApJ, 873, 111, doi: 10.3847/1538-4357/ab042c

    2017 doi

  35. [43]

    M., Nakar, E., Singer, L

    Kasliwal, M. M., Nakar, E., Singer, L. P., et al. 2017a, Science, 358, 1559, doi: 10.1126/science.aap9455

    doi

  36. [44]

    M., Anand, S., Ahumada, T., et al

    Kasliwal, M. M., Anand, S., Ahumada, T., et al. 2020, ApJ, 905, 145, doi: 10.3847/1538-4357/abc335

    2020 doi

  37. [45]

    1986, Astrophysics and Space Science, 118, 5, doi: 10.1007/BF00653622

    Kodaira, K., & Kogure, T. 1986, Astrophysics and Space Science, 118, 5, doi: 10.1007/BF00653622

    1986 doi

  38. [46]

    Kuin, N. P. M., & Team, S. 2019, GRB 190829A: Swift/UVOT Upper Limits, https://gcn.nasa.gov/circulars/24767.txt

    2019

  39. [47]

    M., Gorbovskoy, E., Kornilov, V

    Lipunov, V . M., Gorbovskoy, E., Kornilov, V . G., et al. 2017, ApJL, 850, L1, doi: 10.3847/2041-8213/aa92c0

    2017 doi

  40. [48]

    J., Laher, R

    Masci, F. J., Laher, R. R., Rusholme, B., et al. 2019, PASP, 131, 018003, doi: 10.1088/1538-3873/aae8ac

    2019 doi

  41. [49]

    A., et al

    McCully, C., Hiramatsu, D., Howell, D. A., et al. 2017, ApJL, 848, L32, doi: 10.3847/2041-8213/aa9111

    2017 doi

  42. [50]

    Metzger, B. D. 2017, Living Reviews in Relativity, 20, 3, doi: 10.1007/s41114-017-0006-z

    2017 doi

  43. [51]

    Muthukrishna, D., Parkinson, D., & Tucker, B. E. 2019, ApJ, 885, 85, doi: 10.3847/1538-4357/ab48f4

    2019 doi

  44. [52]

    2017, ApJL, 848, L18, doi: 10.3847/2041-8213/aa9029

    Nicholl, M., Berger, E., Kasen, D., et al. 2017, ApJL, 848, L18, doi: 10.3847/2041-8213/aa9029

    2017 doi

  45. [53]

    2023, Contemporary Physics, 64, 47, doi: 10.1080/00107514.2023.2266921

    Padovani, P., & Cirasuolo, M. 2023, Contemporary Physics, 64, 47, doi: 10.1080/00107514.2023.2266921

    2023

  46. [54]

    Pang, P. T. H., et al. 2023, Nature Communications, 14, 8352, doi: 10.1038/s41467-023-28352-1

    2023 doi

  47. [55]

    P., Coughlin, M

    Petrov, P., Singer, L. P., Coughlin, M. W., et al. 2022, ApJ, 924, 54, doi: 10.3847/1538-4357/ac366d

    2022 doi

  48. [56]

    2017, Nature, 551, 67, doi: 10.1038/nature24298

    Pian, E., D’Avanzo, P., Benetti, S., et al. 2017, Nature, 551, 67, doi: 10.1038/nature24298

    2017 doi

  49. [57]

    S., Barkov, M

    Pozanenko, A. S., Barkov, M. V ., Minaev, P. Y ., et al. 2018, ApJL, 852, L30, doi: 10.3847/2041-8213/aaa2f6

    2018 doi

  50. [58]

    G., & Mohan, S

    Saleem, M., Resmi, L., Arun, K. G., & Mohan, S. 2020, ApJ, 891, 130, doi: 10.3847/1538-4357/ab6731

    2020 doi

  51. [59]

    Sanders, G. H. 2013, Journal of Astrophysics and Astronomy, 34, 81, doi: 10.1007/s12036-013-9169-5

    2013 doi

  52. [60]

    2017, ApJL, 848, L15, doi: 10.3847/2041-8213/aa8f94

    Savchenko, V ., Ferrigno, C., Kuulkers, E., et al. 2017, ApJL, 848, L15, doi: 10.3847/2041-8213/aa8f94

    2017 doi

  53. [61]

    Schuster, M., & Paliwal, K. K. 1997, IEEE Transactions on Signal Processing, 45, 2673, doi: 10.1109/78.650093

    1997 doi

  54. [62]

    J., Simon, J

    Shappee, B. J., Simon, J. D., Drout, M. R., et al. 2017, Science, 358, 1574, doi: 10.1126/science.aaq0186

    2017 doi

  55. [63]

    A., & Johns, M

    Shectman, S. A., & Johns, M. 2003, in Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, V ol. 4837, Large Ground-based Telescopes, ed. J. M. Oschmann & L. M. Stepp, 910–918, doi: 10.1117/12.457909

    2003 doi

  56. [64]

    P., & Price, L

    Singer, L. P., & Price, L. R. 2016, Physical Review D, 93, 024013, doi: 10.1103/PhysRevD.93.024013

    2016 doi

  57. [65]

    J., Chen, T

    Smartt, S. J., Chen, T. W., Jerkstrand, A., et al. 2017, Nature, 551, 75, doi: 10.1038/nature24303

    2017 doi

  58. [66]

    E., Annis, J., et al

    Soares-Santos, M., Holz, D. E., Annis, J., et al. 2017, ApJL, 848, L16, doi: 10.3847/2041-8213/aa9059

    2017 doi

  59. [67]

    2019, ApJL, 881, L40, doi: 10.3847/2041-8213/ab3921 13

    Song, H.-R., Ai, S.-K., Wang, M.-H., et al. 2019, ApJL, 881, L40, doi: 10.3847/2041-8213/ab3921 13

    2019 doi

  60. [68]

    2014, Journal of Machine Learning Research, 15, 1929

    Salakhutdinov, R. 2014, Journal of Machine Learning Research, 15, 1929

    2014

  61. [69]

    R., Levan, A

    Tanvir, N. R., Levan, A. J., Gonz´alez-Fern´andez, C., et al. 2017, ApJL, 848, L27, doi: 10.3847/2041-8213/aa90b6

    2017 doi

  62. [70]

    2024, arXiv e-prints, arXiv:2410.10702, doi: 10.48550/arXiv.2410.10702

    Toivonen, A., Mansingh, G., Griffin, H., et al. 2024, arXiv e-prints, arXiv:2410.10702, doi: 10.48550/arXiv.2410.10702

    2024 doi arXiv

  63. [71]

    2017, PASJ, 69, 101, doi: 10.1093/pasj/psx118

    Utsumi, Y ., Tanaka, M., Tominaga, N., et al. 2017, PASJ, 69, 101, doi: 10.1093/pasj/psx118

    2017 doi

  64. [72]

    J., Yang, S., et al

    Valenti, S., Sand, D. J., Yang, S., et al. 2017, ApJL, 848, L24, doi: 10.3847/2041-8213/aa8edf van der Walt, S., Crellin-Quick, A., & Bloom, J. 2019, The Journal of Open Source Software, 4, 1247, doi: 10.21105/joss.01247

    2017 doi

  65. [73]

    A., Guillochon, J., Berger, E., et al

    Villar, V . A., Guillochon, J., Berger, E., et al. 2017, ApJL, 851, L21, doi: 10.3847/2041-8213/aa9c84

    2017 doi

  66. [74]

    2022, European Physical Journal C, 82, 9, doi: 10.1140/epjc/s10052-021-09981-1 Weizmann Kiendrebeogo, R., Farah, A

    Wang, S., & Zhao, Z.-C. 2022, European Physical Journal C, 82, 9, doi: 10.1140/epjc/s10052-021-09981-1 Weizmann Kiendrebeogo, R., Farah, A. M., Foley, E. M., et al. 2023, arXiv e-prints, arXiv:2306.09234, doi: 10.48550/arXiv.2306.09234 14 APPENDIX A. RUBIN ARCHITECTURE Input F...

    2022 doi arXiv

Pith tools

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