Foundation models for generalist medical artificial intelligence
Open accessThe DOI matches its Crossref record; 70 of the 70 references it cites were found in a registry or an index; 23 of the 33 sources its claims cite were checked for support.
Abstract
The exceptionally rapid development of highly flexible, reusable artificial intelligence (AI) models is likely to usher in newfound capabilities in medicine. We propose a new paradigm for medical AI, which we refer to as generalist medical AI (GMAI). GMAI models will be capable of carrying out a diverse set of tasks using very little or no task-specific labelled data. Built through self-supervision on large, diverse datasets, GMAI will flexibly interpret different combinations of medical modalities, including data from imaging, electronic health records, laboratory results, genomics, graphs or medical text. Models will in turn produce expressive outputs such as free-text explanations, spoken recommendations or image annotations that demonstrate advanced medical reasoning abilities. Here we identify a set of high-impact potential applications for GMAI and lay out specific technical capabilities and training datasets necessary to enable them. We expect that GMAI-enabled applications will challenge current strategies for regulating and validating AI devices for medicine and will shift practices associated with the collection of large medical datasets. This review discusses generalist medical artificial intelligence, identifying potential applications and setting out specific technical capabilities and training datasets necessary to enable them, as well as highlighting challenges to its implementation.
Overview
CoChat Summary
TL;DR
The paper proposes generalist medical AI (GMAI), a class of foundation models that would use self-supervision on large, diverse medical data to perform dynamically specified tasks, flexibly combine modalities, and reason with formal medical knowledge. It outlines six high-impact use cases and argues that GMAI will require new approaches to validation, regulation, bias auditing, privacy, and data collection.
Key findings
- Defines GMAI by three capabilities: dynamic task specification, flexible multimodal inputs/outputs, and formal representation of medical knowledge for advanced reasoning; existing models such as GPT-3 or PaLM show some medical reasoning but are not multimodal and do not reliably generate factual statements.
- Proposes six use cases: grounded radiology reports, augmented procedures, bedside decision support, chatbots for patients, text-to-protein generation, and interactive note-taking.
- Notes current medical AI is narrow: of more than 500 FDA-approved AI models for clinical medicine, most were approved for only 1 or 2 narrow tasks; CheXzero can detect dozens of diseases in chest X-rays without explicit labels; GPT-3 powered more than 300 apps within months of release.
- Flags scale and cost: PaLM, a 540-billion-parameter model, required an estimated 8.4 million hours’ worth of tensor processing unit v4 chips, using roughly 3,000 to 6,000 chips at a time, amounting to millions of dollars; training each model has been estimated to generate up to hundreds of tons of CO₂ equivalent; one study recommended 20 times more tokens than parameters for optimal performance.
Methods
- Perspective/review article; no primary data, sample, or experimental analysis.
- Synthesizes recent foundation-model advances and proposes the GMAI paradigm, identifying technical capabilities, training-data needs, use cases, and challenges.
- Draws on cited prior work and examples (e.g., GPT-3, Gato, CheXzero, PaLM) rather than new experiments.
Limitations
- Stated: GMAI will be uniquely difficult to validate because it can carry out previously unseen tasks, making failure modes hard to anticipate; social bias can increase with model scale; privacy risks include memorization of training data and prompt attacks; scale brings high computational, financial, and environmental costs; verification may require multidisciplinary panels; data collection/access and regulatory/liability frameworks are unresolved.
- Evident from the text: no GMAI model is built or empirically evaluated here; the use cases are prospective and illustrative; many capabilities depend on future advances in multimodal self-supervision, in-context learning, and medical knowledge integration.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Figures


Verification
Claims the paper makes
17 Valid · 3 Not valid · 5 Not checked · 59 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“They can acquire biases during training, when datasets either underrepresent certain groups of patients or contain harmful correlations 44,45 .”
§ Perspective Foundation models for generalist medical artificial intelligence > Social biases.
What the sources say
- Partly supportedcites [44] Obermeyer, Z., Powers, B., Vogeli, C. & Mullainathan, S. Dissecting racial bias in an algorithm used to manage the health of populations. Science 366, 447–453 (2019). sourceExperimental — accuracy not yet measured
The Mechanism of bias section (pp. 4–5) shows that the model learned to predict health care costs, which were correlated with race differently from health needs, producing racial bias. The paper does not identify underrepresentation of patient groups in training data as a cause of the bias it studied. (LLM, conf=0.87)
- Not matchedcites [45] Sex and Gender Bias in Technology and Artificial Intelligence: Biomedicine and Healthcare Applications (Academic, 2022).
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.
- Partly supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Individual models can now achieve state-of-the-art performance on a wide variety of problems, ranging from answering questions about texts to describing images and playing video games [2][3][4] .”
What the sources say
- Not matchedcites [2] Reed, S. et al. A generalist agent. In Transactions on Machine Learning Research (2022). This study presented Gato, a generalist model that can carry out a variety of tasks across modalities such as chatting, captioning images, playing video games and controlling a robot arm.
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Partly supportedcites [3] Alayrac, J.-B. et al. Flamingo: a Visual Language Model for few-shot learning. In Advances in Neural Information Processing Systems (eds Oh, A. H. et al.) 35, 23716–23736 (2022). sourceExperimental — accuracy not yet measured
Section 3.1 (p. 7) reports that Flamingo outperforms previous zero- and few-shot methods across 16 vision-language benchmarks and exceeds fine-tuned state of the art on six tasks. Table 6 (p. 30) includes image scene description, supporting that example. The cited paper does not establish the sentence’s examples of answering questions about texts or playing video games. (LLM, conf=0.86)
- Partly supportedcites [4] Lu, J., Clark, C., Zellers, R., Mottaghi, R. & Kembhavi, A. Unified-IO: a unified model for vision, language, and multi-modal tasks. Preprint at https://arxiv.org/abs/2206.08916 (2022). sourceExperimental — accuracy not yet measured
The Introduction (pp. 1–2) describes one model performing diverse vision and language tasks and reports the top overall score on the GRIT benchmark. The Results (p. 10) report state-of-the-art performance on A-OKVQA and strong image captioning, but say its NLP results lag behind state-of-the-art models. Related Work (p. 11) mentions another model's Atari button presses; it does not establish state-of-the-art video-game performance for Unified-IO. (LLM, conf=0.88)
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.
- Not matched
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Furthermore, building on recent retrieval-based approaches, GMAI can retrieve relevant context from existing databases, in the form of articles, images or entire previous cases 19,20 .”
§ Perspective Foundation models for generalist medical artificial intelligence > Medical domain knowledge
What the sources say
- Partly supportedcites [19] Borgeaud, S. et al. Improving language models by retrieving from trillions of tokens. In Proc. 39th Int. Conference on Machine Learning (eds Chaudhuri, K. et al.) 162, 2206–2240 (2022). sourceExperimental — accuracy not yet measured
The Introduction (p. 1) describes Retro retrieving similar text from a large database, and the training-dataset section (p. 3) includes Wikipedia articles among its text sources. The paper does not establish that GMAI can retrieve images or entire previous medical cases; it supports the retrieval-based approach, but not those extensions. (LLM, conf=0.88)
- Partly supportedcites [20] Guu, K., Lee, K., Tung, Z., Pasupat, P. & Chang, M.-W. REALM: retrieval-augmented language model pre-training. In Proc. 37th Int. Conference on Machine Learning (eds Daumé, H. & Singh, A.) 119, 3929–3938 (2020). sourceExperimental — accuracy not yet measured
The Introduction (p. 1) says REALM retrieves relevant documents from a corpus such as Wikipedia, supporting the article-retrieval part. Future Work (p. 9) proposes image retrieval as a possible extension, not an existing capability. The paper does not establish that GMAI retrieves images or entire previous cases. (LLM, conf=0.88)
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.
- Partly supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Models will be able to solve previously unseen problems simply by having new tasks explained to them (dynamic task specification), without needing to be retrained 3,5 .”
What the sources say
- Partly supportedcites [3] Alayrac, J.-B. et al. Flamingo: a Visual Language Model for few-shot learning. In Advances in Neural Information Processing Systems (eds Oh, A. H. et al.) 35, 23716–23736 (2022). sourceExperimental — accuracy not yet measured
The Introduction says Flamingo is designed to adapt to new vision-language tasks without fine-tuning, and the Contributions passage says it performs multimodal tasks from a few input/output examples. That supports adaptation without retraining, but not the broader claim that GMAI models can solve previously unseen problems simply by having tasks explained to them: Flamingo's demonstrated scope is narrower, and its task specification generally uses examples rather than an explanation alone. (LLM, conf=0.84)
- Not matchedcites [5] Brown, T. et al. Language models are few-shot learners. In Advances in Neural Information Processing Systems (eds Larochelle, H. et al.) 33, 1877–1901 (2020). This study presented the language model GPT-3 and discovered that large language models can carry out in-context learning.
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.
- Partly supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“CheXzero can be applied to detect dozens of diseases in chest X-rays without being trained on explicit labels for these diseases 9 .”
§ Perspective Foundation models for generalist medical artificial intelligence > Applicability.
What the sources say
- Supportedcites [9] Tiu, E. et al. Expert-level detection of pathologies from unannotated chest X-ray images via self-supervised learning. Nat. Biomed. Eng. 6, 1399–1406 (2022). This study demonstrated that CheXzero—an early example of a foundation model in medical AI—can detect diseases on chest X-rays without explicit annotation by learning from natural-language descriptions contained in accompanying clinical reports. source
The Main section describes CheXzero as a zero-shot chest X-ray classifier that needs no explicit pathology labels during training. The Results report an AUC of at least 0.700 for 38 of 57 radiographic findings evaluated on the PadChest test set, supporting the claim that it can detect dozens of conditions. (LLM, conf=0.90)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
32 more claims · 2 with their cited sources checked
Create a free account to see all 84 claims and every source verdict.
- Probably not a substantial claim. 35
- It may state two separate claims as one. 25
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“First, these models are predominantly pretrained using vision-only modeling57,59,60, which neglects not only rich supervisory signals found in pathology reports but also precludes multimodal capabilities such as zero-shot visual-language understanding and cross-model retrieval—a fundamental hallmark in foundation models84,85.”
From A multimodal whole-slide foundation model for pathology (2025) · cites this paper as [85]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Barriers to the clinical implementation of LLMs include their susceptibility to hallucinations, the absence of reliable references [13].”
From Diagnostic performance of Prof. Valmed, ChatGPT-5 Thinking, and OpenEvidence in rheumatology: A comparative evaluation (2026) · cites this paper as [13]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“The emergence of (medical) foundational models (Moor et al., 2023), which are generally multimodal, is expected to accelerate research in optimal data fusion further.”
From Navigating the landscape of multimodal AI in medicine: A scoping review on technical challenges and clinical applications (2025) · cites this paper as Moor et al., 2023
The cited source says what the claim attributes to it.
In the citing paper
“Moor et al. introduced a paradigm for foundation model application in medical AI, named generalist medical AI (GMAI) [16].”
From Challenges in AI-driven Biomedical Multimodal Data Fusion and Analysis (2025) · cites this paper as [16]
The cited source says what the claim attributes to it.
In the citing paper
“Moor et al. proposed a framework of the GMAI, a foundation model designed to handle most clinical tasks, aiming to provide high-quality medical support to both patients and clinicians [16].”
From Challenges in AI-driven Biomedical Multimodal Data Fusion and Analysis (2025) · cites this paper as [16]
5 more claims about this paper
Create a free account to see every claim other papers make about this one.
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
§ Perspective Foundation models for generalist medical artificial intelligence > Conclusion
“The presented example is inspired by a case report 58 .Image of the fistula in panel c adapted from ref. 58, CC BY 3.0.”
§ Perspective Foundation models for generalist medical artificial intelligence > References
“50. Vegunta, R., Vegunta, R. & Kutti Sridharan, G. Secondary aortoduodenal fistula presenting as gastrointestinal bleeding and fungemia. Cureus 11, e5575 (2019).”
Citations
Cites
Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L., & Polosukhin, I. (2017) · arXiv · doi:10.48550/arxiv.1706.03762 · Cited by 120,404
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2019) · Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers) · doi:10.18653/v1/n19-1423 · Cited by 54,325
Highly accurate protein structure prediction with AlphaFold
Jumper, J., Evans, R., Pritzel, A., Green, T., Figurnov, M., Ronneberger, O., Tunyasuvunakool, K., Bates, R., Žídek, A., Potapenko, A., et al. (2021) · Nature · doi:10.1038/s41586-021-03819-2 · Cited by 43,764 · cited 1×
“There have already been early efforts to develop foundation models for biological sequences 33,34 , including RFdiffusion, which generates proteins on the basis of simple specifications (for example, a binding target) 35 .”
BackgroundHigh-Resolution Image Synthesis with Latent Diffusion Models
Rombach, R., Blattmann, A., Lorenz, D., Esser, P., & Ommer, B. (2022) · 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr52688.2022.01042 · Cited by 20,302 · cited 2×
“For certain use cases (for example, chatbots), GMAI models can be stored on central compute clusters maintained by organizations with deep technical expertise, as DALL-E or GPT-3 are.”
BackgroundLearning Transferable Visual Models From Natural Language Supervision
Radford, A., Kim, J. W., Hallacy, C., Ramesh, A., Goh, G., Agarwal, S., Sastry, G., Askell, A., Mishkin, P., Clark, J., et al. (2021) · arXiv · doi:10.48550/arxiv.2103.00020 · Cited by 15,293 · cited 2×
“These include the rise of multimodal architectures 13 and self-supervised learning techniques 14 that dispense with explicit labels (for example, language modelling 15 and contrastive learning 16 ), as well as the advent of in-context learning capabilities 5 .”
BackgroundSudlow, C., Gallacher, J., Allen, N., Beral, V., Burton, P., Danesh, J., Downey, P., Elliott, P., Green, J., Landray, M., et al. (2015) · PLOS Medicine · doi:10.1371/journal.pmed.1001779 · Cited by 12,728 · cited 1×
“Large-scale data-sharing efforts, such as the MIMIC (Medical Information Mart for Intensive Care) database 55 or the UK Biobank 56 , will play a critical role in GMAI, and they should be extended to underrepresented countries to create larger, richer and more inclusive training datasets.”
BackgroundAdvances in Neural Information Processing Systems 14: Proceedings of the 2001 Conference
(2002) · The MIT Press eBooks · doi:10.7551/mitpress/1120.001.0001 · Cited by 8,890
Dissecting racial bias in an algorithm used to manage the health of populations
Obermeyer, Z., Powers, B., Vogeli, C., & Mullainathan, S. (2019) · Science · doi:10.1126/science.aax2342 · Cited by 7,137 · cited 1×
“They can acquire biases during training, when datasets either under-represent certain groups of patients or contain harmful correlations 44,45 .”
BackgroundUniProt: the universal protein knowledgebase
The UniProt Consortium (2017) · Nucleic Acids Research · doi:10.1093/nar/gkw1099 · Cited by 4,436
Training language models to follow instructions with human feedback
Ouyang, L., Wu, J., Jiang, X., Almeida, D., Wainwright, C. L., Mishkin, P., Zhang, C., Agarwal, S., Slama, K., Ray, A., et al. (2022) · arXiv · doi:10.48550/arxiv.2203.02155 · Cited by 4,208 · cited 1×
“Such feedback can then be used to improve model behaviour, following the example of InstructGPT, a model created by using human feedback to refine GPT-3 through reinforcement learning 41 .”
Method
Research works printed but not linked: 2
- Reed, S. et al. A generalist agent. In Transactions on Machine Learning Research (2022). This study presented Gato, a generalist model that can carry out a variety of tasks across modalities such as chatting, captioning images, playing video games and controlling a robot arm.
- Brown, T. et al. Language models are few-shot learners. In Advances in Neural Information Processing Systems (eds Larochelle, H. et al.) 33, 1877–1901 (2020). This study presented the language model GPT-3 and discovered that large language models can carry out in-context learning.
Cited by
Large language models in medicine
Thirunavukarasu, A. J., Ting, D. S. J., Elangovan, K., Gutierrez, L., Tan, T. F., & Ting, D. S. W. (2023) · Nature Medicine · doi:10.1038/s41591-023-02448-8 · Cited by 3,944
Pang, Z., Lu, Y., Zhou, G., Hui, F., Xu, L., Viau, C., Spigelman, A. F., MacDonald, P. E., Wishart, D. S., Li, S., et al. (2024) · Nucleic Acids Research · doi:10.1093/nar/gkae253 · Cited by 2,291
“For future directions, we will continue to improve metabolome annotations, better integrate with other omics data, and explore new ways to interact with users via generative artificial intelligence technologies ( 70–73 ).”
BackgroundTowards a general-purpose foundation model for computational pathology
Chen, R. J., Ding, T., Lu, M. Y., Williamson, D. F. K., Jaume, G., Song, A. H., Chen, B., Zhang, A., Shao, D., Shaban, M., et al. (2024) · Nature Medicine · doi:10.1038/s41591-024-02857-3 · Cited by 1,560
scGPT: Towards Building a Foundation Model for Single-Cell Multi-omics Using Generative AI
Cui, H., Wang, C., Maan, H., Pang, K., Luo, F., & Wang, B. (2023) · doi:10.1101/2023.04.30.538439 · Cited by 1,694
“One promising approach to address this challenge is the generative pre-training of founda39 tion models [16, 17].”
BackgroundA foundation model for generalizable disease detection from retinal images
Zhou, Y., Chia, M. A., Wagner, S. K., Ayhan, M. S., Williamson, D. J., Struyven, R. R., Liu, T., Xu, M., Lozano, M. G., Woodward-Court, P., et al. (2023) · Nature · doi:10.1038/s41586-023-06555-x · Cited by 1,014
Self-citation
Metadata
- Authors
- Moor, M., Banerjee, O., Abad, Z. S. H., Krumholz, H. M., Leskovec, J., Topol, E. J., & Rajpurkar, P.
- Affiliations
- Stanford University (US)
- Harvard University (US)
- University of Toronto (CA)
- Yale New Haven Hospital (US)
- Scripps Research Institute (US)
- Corresponding author
- Topol, Eric J.
- Rajpurkar, Pranav
- ORCID
- Venue
- Nature
- Published
- 2023-04-12
- Volume
- 616
- Issue
- 7956
- Pages
- 259-265
- ISSN
- 0028-0836 (ISSN-L)
- 1476-4687
- OpenAlex
- W4365143687
- Semantic Scholar
- 9faa2b0e5cb93f20df0555c3c350fab0b2eccf3a
- Cited by
- 2,028
- References
- 50
- Citations by year
- 2026: 753 · 2025: 645 · 2024: 445 · 2023: 181
- FWCI
- 23.78
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- cc-by
- www.springernature.com/gp/researchers/text-and-data-mining (tdm)
- www.springernature.com/gp/researchers/text-and-data-mining (vor)
- Locations
- Nature (publishedVersion, cc-by)
- PubMed (publishedVersion)
- RePEc: Research Papers in Economics (submittedVersion)
- Funders
- National Science Foundation (1934578, 1835598, UL1TR001114, 1918940)
- National Institutes of Health (UL1TR001114)
- Defense Advanced Research Projects Agency (HR00112190039, N660011924033)
- Advanced Research Projects Agency
- Multidisciplinary University Research Initiative
- Wu Tsai Neurosciences Institute, Stanford University
- National Human Genome Research Institute (U54HG010426)
- National Institute of Neurological Disorders and Stroke (R61NS11865)
- National Center for Advancing Translational Sciences (grant UL1TR001114)
- Army Research Office (W911NF-16-1-0342, W911NF-16-1-, W911NF)
- Topics
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Physical Sciences › Computer Science › Artificial Intelligence › Machine Learning in Healthcare
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Multimodal Machine Learning Applications
- Keywords
- foundation models, self-supervised learning, multimodal medical data, medical reasoning, explainable AI, health data collection, artificial intelligence in healthcare
- MeSH terms
- Unsupervised Machine Learning, Artificial Intelligence, Diagnostic Imaging, Humans, Medicine, Genomics, Electronic Health Records, Datasets as Topic
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).
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: Foundation models for generalist medical artificial intelligence. https://app.cochat.ai/papers/10.1038/s41586-023-05881-4
- The paper
- Moor, M., Banerjee, O., Abad, Z. S. H., Krumholz, H. M., Leskovec, J., Topol, E. J., & Rajpurkar, P. (2023). Foundation models for generalist medical artificial intelligence. Nature. https://doi.org/10.1038/s41586-023-05881-4
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.