Synthesizing scientific literature with retrieval-augmented language models. Asai, A., He, J., Shao, R., Shi, W., Singh, A., Chang, J. C., Lo, K., Soldaini, L., Feldman, S., D’Arcy, M., Wadden, D., Latzke, M., Sparks, J., Hwang, J. D., Kishore, V., Tian, M., Ji, P., Liu, S., Tong, H., Wu, B., Xiong, Y., Zettlemoyer, L., Neubig, G., Weld, D. S., Downey, D., Yih, W., Koh, P. W., & Hajishirzi, H. Nature, Nature Publishing Group, February, 2026.
Synthesizing scientific literature with retrieval-augmented language models [link]Paper  doi  abstract   bibtex   
Scientific progress depends on the ability of researchers to synthesize the growing body of literature. Can large language models (LLMs) assist scientists in this task? Here we introduce OpenScholar, a specialized retrieval-augmented language model (LM)1 that answers scientific queries by identifying relevant passages from 45 million open-access papers and synthesizing citation-backed responses. To evaluate OpenScholar, we develop ScholarQABench, the first large-scale multi-domain benchmark for literature search, comprising 2,967 expert-written queries and 208 long-form answers across computer science, physics, neuroscience and biomedicine. Despite being a smaller open model, OpenScholar-8B outperforms GPT-4o by 6.1% and PaperQA2 by 5.5% in correctness on a challenging multi-paper synthesis task from the new ScholarQABench. Although GPT-4o hallucinates citations 78–90% of the time, OpenScholar achieves citation accuracy on par with human experts. OpenScholar’s data store, retriever and self-feedback inference loop improve off-the-shelf LMs: for instance, OpenScholar-GPT-4o improves the correctness of GPT-4o by 12%. In human evaluations, experts preferred OpenScholar-8B and OpenScholar-GPT-4o responses over expert-written ones 51% and 70% of the time, respectively, compared with 32% for GPT-4o. We open-source all artefacts, including our code, models, data store, datasets and a public demo.
@article{asaiSynthesizingScientificLiterature2026,
	title = {Synthesizing scientific literature with retrieval-augmented language models},
	copyright = {2026 The Author(s)},
	issn = {1476-4687},
	url = {https://www.nature.com/articles/s41586-025-10072-4},
	doi = {10.1038/s41586-025-10072-4},
	abstract = {Scientific progress depends on the ability of researchers to synthesize the growing body of literature. Can large language models (LLMs) assist scientists in this task? Here we introduce OpenScholar, a specialized retrieval-augmented language model (LM)1 that answers scientific queries by identifying relevant passages from 45 million open-access papers and synthesizing citation-backed responses. To evaluate OpenScholar, we develop ScholarQABench, the first large-scale multi-domain benchmark for literature search, comprising 2,967 expert-written queries and 208 long-form answers across computer science, physics, neuroscience and biomedicine. Despite being a smaller open model, OpenScholar-8B outperforms GPT-4o by 6.1\% and PaperQA2 by 5.5\% in correctness on a challenging multi-paper synthesis task from the new ScholarQABench. Although GPT-4o hallucinates citations 78–90\% of the time, OpenScholar achieves citation accuracy on par with human experts. OpenScholar’s data store, retriever and self-feedback inference loop improve off-the-shelf LMs: for instance, OpenScholar-GPT-4o improves the correctness of GPT-4o by 12\%. In human evaluations, experts preferred OpenScholar-8B and OpenScholar-GPT-4o responses over expert-written ones 51\% and 70\% of the time, respectively, compared with 32\% for GPT-4o. We open-source all artefacts, including our code, models, data store, datasets and a public demo.},
	language = {en},
	urldate = {2026-02-06},
	journal = {Nature},
	publisher = {Nature Publishing Group},
	author = {Asai, Akari and He, Jacqueline and Shao, Rulin and Shi, Weijia and Singh, Amanpreet and Chang, Joseph Chee and Lo, Kyle and Soldaini, Luca and Feldman, Sergey and D’Arcy, Mike and Wadden, David and Latzke, Matt and Sparks, Jenna and Hwang, Jena D. and Kishore, Varsha and Tian, Minyang and Ji, Pan and Liu, Shengyan and Tong, Hao and Wu, Bohao and Xiong, Yanyu and Zettlemoyer, Luke and Neubig, Graham and Weld, Daniel S. and Downey, Doug and Yih, Wen-tau and Koh, Pang Wei and Hajishirzi, Hannaneh},
	month = feb,
	year = {2026},
	keywords = {Computer science, Information technology},
	pages = {1--7},
}
Downloads: 0
About
Documentation
Keywords
Search
Users
Terms and Conditions of Use
Privacy Policy
Sitemap
© BibBase.org 2021