Large language models surpass human experts in predicting neuroscience results
Gespeichert in:
| Titel: | Large language models surpass human experts in predicting neuroscience results |
|---|---|
| Autoren: | Xiaoliang Luo, Akilles Rechardt, Guangzhi Sun, Kevin K. Nejad, Felipe Yáñez, Bati Yilmaz, Kangjoo Lee, Alexandra O. Cohen, Valentina Borghesani, Anton Pashkov, Daniele Marinazzo, Jonathan Nicholas, Alessandro Salatiello, Ilia Sucholutsky, Pasquale Minervini, Sepehr Razavi, Roberta Rocca, Elkhan Yusifov, Tereza Okalova, Nianlong Gu, Martin Ferianc, Mikail Khona, Kaustubh R. Patil, Pui-Shee Lee, Rui Mata, Nicholas E. Myers, Jennifer K. Bizley, Sebastian Musslick, Isil Poyraz Bilgin, Guiomar Niso, Justin M. Ales, Michael Gaebler, N. Apurva Ratan Murty, Leyla Loued-Khenissi, Anna Behler, Chloe M. Hall, Jessica Dafflon, Sherry Dongqi Bao, Bradley C. Love |
| Weitere Verfasser: | Economic and Social Research Council (UK), Microsoft, Apollo - University of Cambridge Repository, University of St Andrews.School of Psychology and Neuroscience, University of St Andrews.Institute of Behavioural and Neural Sciences |
| Quelle: | Nat Hum Behav Digital.CSIC. Repositorio Institucional del CSIC Consejo Superior de Investigaciones Científicas (CSIC) arXiv Nature human behaviour, vol. 9, no. 2, pp. 305-315 Nature human behaviour 9(2), 305-315 (2025). doi:10.1038/s41562-024-02046-9 NATURE HUMAN BEHAVIOUR Luo, X, Rechardt, A, Sun, G, Nejad, K K, Yáñez, F, Yilmaz, B, Lee, K, Cohen, A O, Borghesani, V, Pashkov, A, Marinazzo, D, Nicholas, J, Salatiello, A, Sucholutsky, I, Minervini, P, Razavi, S, Rocca, R, Yusifov, E, Okalova, T, Gu, N, Ferianc, M, Khona, M, Patil, K R, Lee, P S, Mata, R, Myers, N E, Bizley, J K, Musslick, S, Bilgin, I P, Niso, G, Ales, J M, Gaebler, M, Ratan Murty, N A, Loued-Khenissi, L, Behler, A, Hall, C M, Dafflon, J, Bao, S D & Love, B C 2025, 'Large language models surpass human experts in predicting neuroscience results', Nature Human Behaviour, vol. 9, no. 2, pp. 305-315. https://doi.org/10.1038/s41562-024-02046-9 |
| Publication Status: | Preprint |
| Verlagsinformationen: | Springer Science and Business Media LLC, 2024. |
| Publikationsjahr: | 2024 |
| Schlagwörter: | FOS: Computer and information sciences, 0301 basic medicine, Technology and Engineering, Scientific community, Social Psychology, Computer Science - Artificial Intelligence, Social Sciences, Experimental and Cognitive Psychology, Article, neuroscience, Behavioral Neuroscience, 03 medical and health sciences, Large Language Models, Medicine and Health Sciences, Humans, large language models, Ensure healthy lives and promote well-being for all at all ages, Language, 0303 health sciences, Neurosciences, Brain, DAS, artificial intelligence, Brain/physiology, Brain/diagnostic imaging, Artificial Intelligence (cs.AI), Quantitative Biology - Neurons and Cognition, FOS: Biological sciences, Neurons and Cognition (q-bio.NC), Forecasting, Neuroscience |
| Beschreibung: | Scientific discoveries often hinge on synthesizing decades of research, a task that potentially outstrips human information processing capacities. Large language models (LLMs) offer a solution. LLMs trained on the vast scientific literature could potentially integrate noisy yet interrelated findings to forecast novel results better than human experts. Here, to evaluate this possibility, we created BrainBench, a forward-looking benchmark for predicting neuroscience results. We find that LLMs surpass experts in predicting experimental outcomes. BrainGPT, an LLM we tuned on the neuroscience literature, performed better yet. Like human experts, when LLMs indicated high confidence in their predictions, their responses were more likely to be correct, which presages a future where LLMs assist humans in making discoveries. Our approach is not neuroscience specific and is transferable to other knowledge-intensive endeavours. |
| Publikationsart: | Article Other literature type |
| Dateibeschreibung: | application/pdf; application/zip; text/xml |
| Sprache: | English |
| ISSN: | 2397-3374 |
| DOI: | 10.1038/s41562-024-02046-9 |
| DOI: | 10.48550/arxiv.2403.03230 |
| DOI: | 10.34734/fzj-2025-02220 |
| Zugangs-URL: | https://pubmed.ncbi.nlm.nih.gov/39604572 http://arxiv.org/abs/2403.03230 http://hdl.handle.net/10261/395833 https://api.elsevier.com/content/abstract/scopus_id/85203189734 http://hdl.handle.net/21.11116/0000-000F-68CD-6 http://hdl.handle.net/21.11116/0000-000F-68CF-4 http://hdl.handle.net/21.11116/0000-000F-68D1-0 https://serval.unil.ch/notice/serval:BIB_FF43875B71A5 https://serval.unil.ch/resource/serval:BIB_FF43875B71A5.P001/REF.pdf http://nbn-resolving.org/urn/resolver.pl?urn=urn:nbn:ch:serval-BIB_FF43875B71A54 https://juser.fz-juelich.de/record/1041321 http://hdl.handle.net/1854/LU-01JDRY0F05A02KS70NW5GRW0DB http://doi.org/10.1038/s41562-024-02046-9 https://biblio.ugent.be/publication/01JDRY0F05A02KS70NW5GRW0DB https://biblio.ugent.be/publication/01JDRY0F05A02KS70NW5GRW0DB/file/01JDRY4W5S0BFNHY0JQ4NC8X07 http://www.scopus.com/inward/record.url?scp=85203189734&partnerID=8YFLogxK https://hdl.handle.net/10023/31983 |
| Rights: | CC BY |
| Dokumentencode: | edsair.doi.dedup.....af923d6f99a44c7e227f4ec4bb986c9b |
| Datenbank: | OpenAIRE |
Full Text Finder 
Nájsť tento článok vo Web of Science | Abstract: | Scientific discoveries often hinge on synthesizing decades of research, a task that potentially outstrips human information processing capacities. Large language models (LLMs) offer a solution. LLMs trained on the vast scientific literature could potentially integrate noisy yet interrelated findings to forecast novel results better than human experts. Here, to evaluate this possibility, we created BrainBench, a forward-looking benchmark for predicting neuroscience results. We find that LLMs surpass experts in predicting experimental outcomes. BrainGPT, an LLM we tuned on the neuroscience literature, performed better yet. Like human experts, when LLMs indicated high confidence in their predictions, their responses were more likely to be correct, which presages a future where LLMs assist humans in making discoveries. Our approach is not neuroscience specific and is transferable to other knowledge-intensive endeavours. |
|---|---|
| ISSN: | 23973374 |
| DOI: | 10.1038/s41562-024-02046-9 |