EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Generalization bias in large language model summarization of scientific research

Uwe Peters iD, Benjamin Chin-Yee iD

Royal Society Open Science · 2025-04 · 12 (4)

DOI: 10.1098/rsos.241776 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Royal Society Open Science, online erschienen am 30.04.2025. Verfasst von Uwe Peters und 1 weiteren Autor:innen. Beteiligte Einrichtung: Utrecht University, Utrecht, The Netherlands. Die Arbeit wurde bisher 141-mal zitiert (Crossref) und nennt 59 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Artificial intelligence chatbots driven by large language models (LLMs) have the potential to increase public science literacy and support scientific research, as they can quickly summarize complex scientific information in accessible terms. However, when summarizing scientific texts, LLMs may omit details that limit the scope of research conclusions, leading to generalizations of results broader than warranted by the original study. We tested 10 prominent LLMs, including ChatGPT-4o, ChatGPT-4.5, DeepSeek, LLaMA 3.3 70B, and Claude 3.7 Sonnet, comparing 4900 LLM-generated summaries to their original scientific texts. Even when explicitly prompted for accuracy, most LLMs produced broader generalizations of scientific results than those in the original texts, with DeepSeek, ChatGPT-4o, and LLaMA 3.3 70B overgeneralizing in 26–73% of cases. In a direct comparison of LLM-generated and human-authored science summaries, LLM summaries were nearly five times more likely to contain broad generalizations (odds ratio = 4.85, 95% CI [3.06, 7.70], p < 0.001). Notably, newer models tended to perform worse in generalization accuracy than earlier ones. Our results indicate a strong bias in many widely used LLMs towards overgeneralizing scientific conclusions, posing a significant risk of large-scale misinterpretations of research findings. We highlight potential mitigation strategies, including lowering LLM temperature settings and benchmarking LLMs for generalization accuracy.

Schlagwörter

Stichwörter der Autor:innen

algorithmic biaslarge language modelsovergeneralizationscience communication

Autor:innen und Institutionen

  • Utrecht University, Utrecht, The Netherlands.
  • Western University, London, Canada.
  • University of Cambridge, Cambridge, UK.

ORCID-Profile: Uwe Peters, Benjamin Chin-Yee

Zitieren

APA

Peters, U., & Chin-Yee, B. (2025). Generalization bias in large language model summarization of scientific research. Royal Society Open Science, 12(4). https://doi.org/10.1098/rsos.241776

Vancouver

Peters U, Chin-Yee B. Generalization bias in large language model summarization of scientific research. Royal Society Open Science. 2025;12(4). doi:10.1098/rsos.241776

BibTeX
@article{Peters2025,
  author = {Peters, Uwe and Chin-Yee, Benjamin},
  title = {Generalization bias in large language model summarization of scientific research},
  journal = {Royal Society Open Science},
  year = {2025},
  volume = {12},
  number = {4},
  doi = {10.1098/rsos.241776}
}
Kostenlos starten →