EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

Reducing hallucinations of large language models via hierarchical semantic piece

Yanyi Liu, Qingwen Yang, Jiawei Tang, Tiezheng Guo, Chen Wang, Pan Li, Sai Xu, Xianlin Gao, Zhi Li, Jun Liu, Yingyou Wen

Complex & Intelligent Systems · 2025-05 · 11 (5)

DOI: 10.1007/s40747-025-01833-9 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Complex & Intelligent Systems, online erschienen am 02.04.2025. Verfasst von Yanyi Liu und 10 weiteren Autor:innen. Die Arbeit wurde bisher 20-mal zitiert (Crossref) und nennt 58 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: Liaoning Provincial Science and Technology Innovation Project in the Field of Artificial Intelligence, Shenyang Young and Middle-aged Science and Technology Innovation Talent Support Program.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

With the widespread application of large language models (LLMs) in natural language processing (NLP), hallucinations have become a significant impediment to their effective use of LLMs in industry applications. To address this challenge, we integrate existing hallucination detection and mitigation methods into a unified hallucination detection and mitigation framework. The framework consists of four main components: output parser, reference parser, fact verifier, and mitigator. These components collectively consolidate various hallucination detection and mitigation methods. Within this unified framework, we introduce the hierarchical semantic piece (HSP) for hallucination detection and mitigation. The HSP method extracts multi-granularity semantic pieces from both the reference material and the generated text. Sentence-level semantic pieces encapsulate global semantic information, while entity-level semantic pieces handle local semantic information. This method verifies the consistency between the generated text and the reference text at corresponding granularities, thereby enhancing the effectiveness of hallucination detection and mitigation. Experimental results show that the HSP method is very effective in detecting and mitigating hallucinations and shows lower computational resource consumption. Our method has great potential and promises for industry applications that rely on professionalism and reliability.

Förderung

  • Liaoning Provincial Science and Technology Innovation Project in the Field of Artificial Intelligence
  • Shenyang Young and Middle-aged Science and Technology Innovation Talent Support Program

Zitieren

APA

Liu, Y., Yang, Q., Tang, J., Guo, T., Wang, C., Li, P., Xu, S., Gao, X., Li, Z., Liu, J., & Wen, Y. (2025). Reducing hallucinations of large language models via hierarchical semantic piece. Complex & Intelligent Systems, 11(5). https://doi.org/10.1007/s40747-025-01833-9

Vancouver

Liu Y, Yang Q, Tang J, Guo T, Wang C, Li P, et al. Reducing hallucinations of large language models via hierarchical semantic piece. Complex & Intelligent Systems. 2025;11(5). doi:10.1007/s40747-025-01833-9

BibTeX
@article{Liu2025,
  author = {Liu, Yanyi and Yang, Qingwen and Tang, Jiawei and Guo, Tiezheng and Wang, Chen and Li, Pan and Xu, Sai and Gao, Xianlin and Li, Zhi and Liu, Jun and Wen, Yingyou},
  title = {Reducing hallucinations of large language models via hierarchical semantic piece},
  journal = {Complex & Intelligent Systems},
  year = {2025},
  volume = {11},
  number = {5},
  doi = {10.1007/s40747-025-01833-9}
}
Kostenlos starten →