AI hallucinations can’t be stopped — but these techniques can limit their damage
Author:
- Nicola JonesNJNicola Jones
Top Cited Papers
- 21 January 2025
- journal article
- news
- Published by Springer Nature in Nature
- Vol. 637 (8047) , 778-780
Abstract
No abstract availableKeywords
This publication has 4 references indexed in Scilit:
- Larger and more instructable language models become less reliableNature, 2024
- Detecting hallucinations in large language models using semantic entropyNature, 2024
- ChatGPT is bullshitEthics and Information Technology, 2024
- Hallucination Rates and Reference Accuracy of ChatGPT and Bard for Systematic Reviews: Comparative AnalysisJournal of Medical Internet Research, 2024