Kurz gefasst
Fachartikel in der Zeitschrift Queue, online erschienen am 09.09.2024. Verfasst von Jim Waldo und 1 weiteren Autor:innen. Die Arbeit wurde bisher 31-mal zitiert (Crossref) und nennt 7 Literaturangaben.
Abstract (Originalsprache)
The findings in this experiment support the hypothesis that GPTs based on LLMs perform well on prompts that are more popular and have reached a general consensus yet struggle on controversial topics or topics with limited data. The variability in the applications's responses underscores that the models depend on the quantity and quality of their training data, paralleling the system of crowdsourcing that relies on diverse and credible contributions. Thus, while GPTs can serve as useful tools for many mundane tasks, their engagement with obscure and polarized topics should be interpreted with caution. LLMs' reliance on probabilistic models to produce statements about the world ties their accuracy closely to the breadth and quality of the data they're given.
Zitieren
APA
Waldo, J., & Boussard, S. (2024). GPTs and Hallucination. Queue, 22(4), 19-33. https://doi.org/10.1145/3688007
Vancouver
Waldo J, Boussard S. GPTs and Hallucination. Queue. 2024;22(4):19-33. doi:10.1145/3688007
BibTeX
@article{Waldo2024,
author = {Waldo, Jim and Boussard, Soline},
title = {GPTs and Hallucination},
journal = {Queue},
year = {2024},
volume = {22},
number = {4},
pages = {19--33},
doi = {10.1145/3688007}
}
EUVIMED