The DOI matches its Crossref record; 125 of the 138 references it cites were found in a registry or an index; 41 of the 79 sources its claims cite were checked for support.
Kraken reveals itself – the merger history of the Milky Way reconstructed with the E-MOSAICS simulations
Open accessAbstract
Globular clusters (GCs) formed when the Milky Way experienced a phase of rapid assembly. We use the wealth of information contained in the Galactic GC population to quantify the properties of the satellite galaxies from which the Milky Way assembled. To achieve this, we train an artificial neural network on the E-MOSAICS cosmological simulations of the co-formation and co-evolution of GCs and their host galaxies. The network uses the ages, metallicities, and orbital properties of GCs that formed in the same progenitor galaxies to predict the stellar masses and accretion redshifts of these progenitors. We apply the network to Galactic GCs associated with five progenitors: Gaia-Enceladus, the Helmi streams, Sequoia, Sagittarius, and the recently discovered ‘low-energy’ GCs, which provide an excellent match to the predicted properties of the enigmatic galaxy ‘Kraken’. The five galaxies cover a narrow stellar mass range [M⋆ = (0.6–4.6) × 108 M⊙], but have widely different accretion redshifts ($\mbox{$z_{\rm acc}$}=0.57\!-\!2.65$). All accretion events represent minor mergers, but Kraken likely represents the most major merger ever experienced by the Milky Way, with stellar and virial mass ratios of $\mbox{$r_{M_\star }$}=1$:$31^{+34}_{-16}$ and $\mbox{$r_{M_{\rm h}}$}=1$:$7^{+4}_{-2}$, respectively. The progenitors match the z = 0 relation between GC number and halo virial mass, but have elevated specific frequencies, suggesting an evolution with redshift. Even though these progenitors likely were the Milky Way’s most massive accretion events, they contributed a total mass of only log (M⋆, tot/M⊙) = 9.0 ± 0.1, similar to the stellar halo. This implies that the Milky Way grew its stellar mass mostly by in-situ star formation. We conclude by organizing these accretion events into the most detailed reconstruction to date of the Milky Way’s merger tree.
Overview
CoChat Summary
TL;DR
Kruijssen et al. train an artificial neural network on the E-MOSAICS cosmological simulations to infer the stellar masses and accretion redshifts of five Milky Way satellite progenitors from globular cluster ages, metallicities, and orbital properties. They identify the low-energy GC group as Kraken, find that all five accretion events were minor mergers, and reconstruct the Milky Way’s merger tree.
Key findings
- The neural network predicts satellite stellar masses with scatter \(\sigma[\log_{10}(M_{\star,\mathrm{pred}}/M_{\star,\mathrm{true}})] = 0.41^{+0.05}_{-0.04}\) and accretion redshifts with \(\sigma[\log_{10}(1+z_{\mathrm{acc,pred}}/1+z_{\mathrm{acc,true}})] = 0.13^{+0.01}_{-0.01}\), with validation score \(0.89^{+0.06}_{-0.06}\).
- The five progenitors span \(M_{\star} = (0.6\text{–}4.6)\times 10^8\,M_\odot\) and \(z_{\mathrm{acc}} = 0.57\text{–}2.65\). Kraken has \(M_{\star}=1.9^{+1.0}_{-0.6}\times 10^8\,M_\odot\) and \(z_{\mathrm{acc}}=2.26^{+0.39}_{-0.45}\); Gaia-Enceladus has \(M_{\star}=2.7^{+1.1}_{-0.8}\times 10^8\,M_\odot\) and \(z_{\mathrm{acc}}=1.35^{+0.26}_{-0.23}\); Sagittarius has \(M_{\star}=2.8^{+1.8}_{-1.1}\times 10^8\,M_\odot\) and \(z_{\mathrm{acc}}=0.76^{+0.22}_{-0.19}\).
- All accretion events are minor mergers. Kraken is likely the most major merger ever experienced by the Milky Way, with stellar mass ratio \(r_{M_\star}=1{:}31^{+34}_{-16}\) and halo mass ratio \(r_{M_h}=1{:}7^{+4}_{-2}\); its probability of being a major merger is 0.3 per cent.
- The five progenitors contributed total stellar mass \(\log(M_{\star,\mathrm{tot}}/M_\odot)=9.0\pm0.1\), similar to the stellar halo, and total halo mass \(\log(M_{h,\mathrm{tot}}/M_\odot)=11.57\pm0.04\), or 25–45 per cent of the Milky Way’s virial mass. The paper estimates 35–50 per cent of Galactic GCs formed ex-situ.
Methods
- The authors train a sequential artificial neural network on 25 E-MOSAICS zoom-in cosmological simulations, using 205 accretion events with \(\log(M_\star/M_\odot)>6.5\) that host GCs. The eight feature variables are the medians and interquartile ranges of GC age, metallicity, orbital apocentre radius, and eccentricity; the two target variables are accretion redshift and host stellar mass at accretion.
- They run 10,000 Monte Carlo realisations with scikit-learn and Keras, using dense hidden layers, ReLU activation, Adam optimizer, mean squared error loss, and early stopping. The network is applied to Galactic GCs associated with five progenitors following Massari et al. (2019), with small changes, using ages and metallicities from Kruijssen et al. (2019b) and orbits from Baumgardt et al. (2019).
- They consider all GC membership permutations, combine predictions with the Milky Way stellar mass growth history from Kruijssen et al. (2019b), and use the Moster et al. (2013) stellar-mass–halo-mass relation to infer halo masses and merger halo mass ratios.
Limitations
- Stated: E-MOSAICS provides only 25 Milky Way-mass galaxies, and the Milky Way’s accretion history is atypical, so the training sample is small; the authors note they do not find strong variations across the suite but cannot guarantee the Milky Way’s particular history is captured.
- Stated: quoted error bars reflect random uncertainties; comparison to the network scatter suggests additional systematic uncertainties of up to 0.3 dex in stellar mass and up to 0.25 in redshift. GC membership ambiguities affect Sequoia and the Helmi streams more strongly, and GC counts are lower limits.
- Evident from the text: the method depends on the E-MOSAICS sub-grid GC formation/disruption model and on the simulation definition of accretion, and the application is to a single galaxy, so broader generality remains to be tested with larger simulation volumes.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Figures









Verification
Claims the paper makes
29 Valid · 6 Not valid · 29 Not checked · 152 Own results and interpretations
Some cited sources support this claim and some don’t — read each source’s verdict below.
In the paper
“suggesting that the scaling relations between the number of GCs and the host galaxy mass evolve with redshift (e.g. Kruijssen 2015; Choksi & Gnedin 2019, Bastian et al. in prep.).”
What the sources say
- Disputedcites Kruijssen 2015 sourceExperimental — accuracy not yet measured
Section 4.3 (pp. 14–15) says the relation between GC number per unit stellar mass and host-galaxy mass is established by the end of the rapid-disruption phase and evolves very little afterward, except for the most metal-rich GCs. Section 5.2 (p. 18) likewise predicts that relations between GC number and host-galaxy properties were already in place at redshift 1–2. The paper therefore predicts little subsequent evolution, rather than a generally evolving scaling relation. (LLM, conf=0.83)
Checked by a reasoning model — may be wrong.
- Partly supportedcites Choksi & Gnedin 2019 sourceExperimental — accuracy not yet measured
Our copy does establish redshift evolution of a GC-host scaling relation: the ABSTRACT (p. 1) says the ratio M_GC/M_h 'is nearly constant at all redshifts, but its normalization evolves by a factor of 10 from birth to z = 0', and Section 3 (pp. 3-4) plus Summary item (i) (p. 10) give the same factor-of-ten decrease between z~3 and z=0, explicitly contrasted with Kruijssen (2015)'s no-evolution prediction (Sec. 6.1, p. 9). However, the quantity shown to evolve is the COMBINED MASS in GCs versus the host DARK MATTER HALO mass (and its slope/shape is said to be largely preserved, only the normalization evolving); the paper's number-based relation, the specific frequency N_GC/M_*, is treated only at z = 0 (Fig. 2, Sec. 3 p. 4; Summary item (ii) p. 10). So the part of the claim attributing the evolution to the NUMBER of GCs against host galaxy mass is not stated in this source. (LLM, conf=0.72)
- Not matchedcites Bastian et al. in prep.
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
hedgedstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Disputed
At least one cited source doesn’t support this claim or points to a different paper, and none confirmed it — see the verdicts below.
In the paper
“For these reasons, the resulting estimates of the stellar masses represent (quite uncertain) upper limits, because they e ectively represent projected masses at z = 0 rather than at the time of accretion.”
What the sources say
- Disputedcites Forbes (2020) sourceExperimental — accuracy not yet measured
Section 6 (p. 6) says Forbes used a near-zero-redshift relation to estimate stellar masses, but explicitly says its evolution is minor for low-mass galaxies and the predicted masses are not particularly sensitive to the merger epoch. The paper instead identifies the inferred metallicities—not the stellar masses—as upper limits at infall (section 6, p. 6; Conclusions, p. 8). (LLM, conf=0.85)
Checked by a reasoning model — may be wrong.
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Disputed
At least one cited source doesn’t support this claim or points to a different paper, and none confirmed it — see the verdicts below.
In the paper
“In part, the discrepancy arises because Helmi et al. (2018) assume that Gaia -Enceladus was similar in stellar mass to the Large Magellanic Cloud, which we find is an overestimation.”
What the sources say
- Disputedcites Helmi et al. (2018) sourceExperimental — accuracy not yet measured
The main text (p. 3) estimates Gaia-Enceladus’s stellar mass at about 6 × 10^8 solar masses, comparable to the present-day Small Magellanic Cloud—not the Large Magellanic Cloud. Its comparison with the Large Magellanic Cloud concerns chemical abundances and the Cloud’s early years, not similar stellar mass. (LLM, conf=0.90)
Checked by a reasoning model — may be wrong.
plainstates a possible cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Disputed
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Because E-MOSAICS overpredicts the number of GCs at masses much smaller than M 10 5 M due to underdisruption (Pfe er et al. 2018; Kruijssen et al. 2019a), we only consider GCs with z = 0 masses of M > 5 10 4 M .”
What the sources say
- Supportedcites Pfeffer et al. 2018 source
Section 6.2 (pp. 25–26) reports too many low-mass clusters in E-MOSAICS and attributes the excess to insufficient disruption by tidal shocks. The mass cut at z = 0 is the citing authors’ stated choice, not a finding attributed to Pfeffer et al. (LLM, conf=0.92)
- Not matchedcites Kruijssen et al. 2019a
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“Relative to other measurements in the literature, we find a lower mass for Gaia -Enceladus than Helmi et al. (2018), who provide a very rough estimate of M ? 6 10 8 M from the star formation rate ( 0 : 3 M yr 1 ) and duration ( 2 Gyr) necessary to reproduce the -poor stellar population (Fern´ andez-Alvar et al. 2018).”
What the sources say
- Supportedcites Helmi et al. (2018) source
The main text (p. 3) estimates a Gaia-Enceladus progenitor stellar mass of about 6 × 10^8 solar masses from a star formation rate of 0.3 solar masses per year lasting about 2 Gyr. The citing paper’s lower mass is its own finding, not a finding attributed to Helmi et al. (LLM, conf=0.94)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
130 more claims · 11 with their cited sources checked
Create a free account to see all 216 claims and to verify the 4 references still open.
- Probably not a substantial claim. 16
- It may state two separate claims as one. 66
- The quote only partly states this claim. 3
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ 3.2.2 Accretion redshifts · p. 9
“For all five satellite progenitors the lookback time of accretion is consistent with (or larger than) the age of the youngest associated GC, which provides an important consistency check (see Figure 3).”
§ Kraken reveals itself 5 · p. 6
“Finally, we see no GCs with ages younger than the accretion redshifts inferred in Section 3.2, which is an important consistency check.”
Citations
Cites
Scikit-learn: Machine Learning in Python
Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Müller, A., Nothman, J., Louppe, G., et al. (2012) · arXiv · doi:10.48550/arxiv.1201.0490 · Cited by 71,016 · cited 1×
“…neural network to predict the accretion redshifts and stellar masses of galaxies accreted by the Milky Way using the Python packages scikit-learn (Pedregosa et al. 2011; Buitinck et al. 2013) and Keras (Chollet et al. 2015), which are application programming interfaces for machine learning and…”
MethodMatplotlib: a 2D graphics environment
Hunter, J. D. (2007) · Computing in Science & Engineering · doi:10.1109/mcse.2007.55 · Cited by 39,249 · cited 1×
“This work has made use of Matplotlib (Hunter 2007), Numpy (van der Walt et al. 2011), and Astropy (Astropy Collaboration et al. 2013).”
MethodAstropy: A community Python package for astronomy
The Astropy Collaboration, Robitaille, T. P., Tollerud, E. J., Greenfield, P., Droettboom, M., Bray, E., Aldcroft, T., Davis, M., Ginsburg, A., Price-Whelan, A. M., et al. (2013) · Astronomy & Astrophysics · doi:10.1051/0004-6361/201322068 · Cited by 14,620
The NumPy Array: A Structure for Efficient Numerical Computation
van der Walt, S., Colbert, S. C., & Varoquaux, G. (2011) · Computing in Science & Engineering · doi:10.1109/mcse.2011.37 · Cited by 9,381 · cited 1×
“This work has made use of Matplotlib (Hunter 2007), Numpy (van der Walt et al. 2011), and Astropy (Astropy Collaboration et al. 2013).”
MethodA Catalog of Parameters for Globular Clusters in the Milky Way
Harris, W. E. (1996) · The Astronomical Journal · doi:10.1086/118116 · Cited by 4,500
The EAGLE project: simulating the evolution and assembly of galaxies and their environments
Schaye, J., Crain, R. A., Bower, R. G., Furlong, M., Schaller, M., Theuns, T., Dalla Vecchia, C., Frenk, C. S., McCarthy, I. G., Helly, J. C., et al. (2015) · Monthly Notices of the Royal Astronomical Society · doi:10.1093/mnras/stu2058 · Cited by 3,601 · cited 1×
“This suite of simulations combines the model for galaxy formation and evolution from EAGLE (Crain et al. 2015; Schaye et al. 2015) with the sub-grid model for the formation and evolution of the entire stellar cluster population MOSAICS (Kruijssen et al. 2011, 2012; Pfeffer et al. 2018).”
MethodSelf-citationTHE AVERAGE STAR FORMATION HISTORIES OF GALAXIES IN DARK MATTER HALOS FROM z = 0-8
Behroozi, P. S., Wechsler, R. H., & Conroy, C. (2013) · The Astrophysical Journal · doi:10.1088/0004-637x/770/1/57 · Cited by 2,167 · cited 1×
“This is consistent with the abundance matching models of Moster et al. (2013), who predict a stellar mass growth by a factor of ∼ 5 over the same time interval, and of Behroozi et al. (2013), who predict a factor of ∼ 4.”
ResultPlanck 2013 results. I. Overview of products and scientific results
Planck Collaboration, Ade, P. A. R., Aghanim, N., Alves, M. I. R., Armitage-Caplan, C., Arnaud, M., Ashdown, M., Atrio-Barandela, F., Aumont, J., Aussel, H., et al. (2014) · Astronomy & Astrophysics · doi:10.1051/0004-6361/201321529 · Cited by 1,738
Evidence from the motions of old stars that the Galaxy collapsed.
Eggen, O. J., Lynden-Bell, D., & Sandage, A. R. (1962) · The Astrophysical Journal · doi:10.1086/147433 · Cited by 1,687 · cited 1×
“It is one of the main goals in modern galaxy formation studies to reconstruct and understand the assembly histories of galaxies (e.g. Eggen et al. 1962; Searle & Zinn 1978; Ibata et al. 1994; Belokurov et al. 2006; Bell et al. 2008; Johnston et al. 2008; McConnachie et al. 2009; Cooper et al. 2010;…”
BackgroundGADGET: a code for collisionless and gasdynamical cosmological simulations
Springel, V., Yoshida, N., & White, S. D. M. (2001) · New Astronomy · doi:10.1016/s1384-1076(01)00042-2 · Cited by 1,664 · cited 1×
“In E-MOSAICS, the accretion redshift is defined as the moment at which subfind (Springel et al. 2001; Dolag et al. 2009) can no longer find a bound subhalo and the subhalo is therefore considered to have merged into the halo of the central halo (see Qu et al. 2017 for discussion).”
Background
Research works printed but not linked: 8
- Bastian, 2020, MNRAS
- Bonaca, 2020, ApJ
- Buitinck, 2013, ECML PKDD Workshop: Languages for Data Mining and Machine Learning, 108
- Chollet, 2015, Keras
- Horta, 2020, MNRAS
- Pedregosa, 2011, J. Mach. Learn. Res., 12, 2825
- Pfeffer, 2020, MNRAS
- Trujillo-Gomez, 2020, MNRAS
Cited by
Gaia EDR3 view on galactic globular clusters
Vasiliev, E., & Baumgardt, H. (2021) · Monthly Notices of the Royal Astronomical Society · doi:10.1093/mnras/stab1475 · Cited by 484
Anders, F., Khalatyan, A., Queiroz, A. B. A., Chiappini, C., Ardèvol, J., Casamiquela, L., Figueras, F., Jiménez-Arranz, Ó., Jordi, C., Monguió, M., et al. (2022) · Astronomy & Astrophysics · doi:10.1051/0004-6361/202142369 · Cited by 240
Horta, D., Schiavon, R. P., Mackereth, J. T., Pfeffer, J., Mason, A. C., Kisku, S., Fragkoudi, F., Allende Prieto, C., Cunha, K., Hasselquist, S., et al. (2020) · Monthly Notices of the Royal Astronomical Society · doi:10.1093/mnras/staa2987 · Cited by 219
“Indeed, both the orbital properties and time of accretion of the progenitor of the IGS agree with the predictions by Kruijssen et al. (2019b, 2020) for the Kraken accretion event.”
BackgroundResultInfluentialSelf-citationMalhan, K., Ibata, R. A., Sharma, S., Famaey, B., Bellazzini, M., Carlberg, R. G., D’Souza, R., Yuan, Z., Martin, N. F., & Thomas, G. F. (2022) · The Astrophysical Journal · doi:10.3847/1538-4357/ac4d2a · Cited by 171
“Among our member objects, NGC 6441 was tentatively associated with the Kraken merger by Kruijssen et al. (2020).”
BackgroundMethodResultInfluentialReconstructing the Last Major Merger of the Milky Way with the H3 Survey
Naidu, R. P., Conroy, C., Bonaca, A., Zaritsky, D., Weinberger, R., Ting 丁, Y.-S. 源., Caldwell, N., Tacchella, S., Han, J. J., Speagle, J. S., et al. (2021) · The Astrophysical Journal · doi:10.3847/1538-4357/ac2d2d · Cited by 163
“Note that this is 5–10× less massive than estimated in Myeong et al. (2019) and Kruijssen et al. (2020).”
BackgroundInfluentialSelf-citation
Metadata
- Authors
- Kruijssen, J. M. D., Pfeffer, J. L., Chevance, M., Bonaca, A., Trujillo-Gomez, S., Bastian, N., Reina-Campos, M., Crain, R. A., & Hughes, M. E.
- Affiliations
- Harvard University (US)
- Liverpool John Moores University (GB)
- Heidelberg University (DE)
- Corresponding author
- Kruijssen, J M Diederik
- ORCID
- Venue
- Monthly Notices of the Royal Astronomical Society
- Publisher
- Oxford University Press (OUP)
- Published
- 2020-08-17
- Volume
- 498
- Issue
- 2
- Pages
- 2472-2491
- ISSN
- 0035-8711 (ISSN-L)
- 1365-2966
- OpenAlex
- W3009673653
- Cited by
- 252
- References
- 138
- Citations by year
- 2026: 25 · 2025: 32 · 2024: 42 · 2023: 56 · 2022: 41 · 2021: 37 · 2020: 19
- FWCI
- 17.05
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Licence
- academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model (vor)
- Locations
- Monthly Notices of the Royal Astronomical Society (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- Liverpool John Moores University (submittedVersion)
- arXiv (Cornell University)
- Funders
- California Department of Fish and Game (KR4801/1-1, KR4801/2-1)
- H2020 European Research Council (714907)
- Topics
- Physical Sciences › Physics and Astronomy › Astronomy and Astrophysics › Galaxies: Formation, Evolution, Phenomena
- Physical Sciences › Physics and Astronomy › Astronomy and Astrophysics › Stellar, planetary, and galactic studies
- Physical Sciences › Physics and Astronomy › Astronomy and Astrophysics › Astrophysical Phenomena and Observations
- Keywords
- globular clusters, satellite galaxies, Gaia-Enceladus, Helmi streams, Sagittarius dwarf galaxy, Sequoia, minor mergers, major mergers, stellar halo, in situ star formation, artificial neural networks
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol and qwen3.8-omni-flash).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Kraken reveals itself – the merger history of the Milky Way reconstructed with the E-MOSAICS simulations. https://app.cochat.ai/papers/10.1093/mnras/staa2452
- The paper
- Kruijssen, J. M. D., Pfeffer, J. L., Chevance, M., Bonaca, A., Trujillo-Gomez, S., Bastian, N., Reina-Campos, M., Crain, R. A., & Hughes, M. E. (2020). Kraken reveals itself – the merger history of the Milky Way reconstructed with the E-MOSAICS simulations. Monthly Notices of the Royal Astronomical Society. https://doi.org/10.1093/mnras/staa2452
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.