Guardado en:
Detalles Bibliográficos
Autores principales: Selby, David, Spriestersbach, Kai, Iwashita, Yuichiro, Saad, Mohammad, Bappert, Dennis, Warrier, Archana, Mukherjee, Sumantrak, Kise, Koichi, Vollmer, Sebastian
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:https://arxiv.org/abs/2402.07770
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866913743897624576
author Selby, David
Spriestersbach, Kai
Iwashita, Yuichiro
Saad, Mohammad
Bappert, Dennis
Warrier, Archana
Mukherjee, Sumantrak
Kise, Koichi
Vollmer, Sebastian
author_facet Selby, David
Spriestersbach, Kai
Iwashita, Yuichiro
Saad, Mohammad
Bappert, Dennis
Warrier, Archana
Mukherjee, Sumantrak
Kise, Koichi
Vollmer, Sebastian
contents Large language models (LLMs) have been extensively studied for their abilities to generate convincing natural language sequences, however their utility for quantitative information retrieval is less well understood. Here we explore the feasibility of LLMs as a mechanism for quantitative knowledge retrieval to aid two data analysis tasks: elicitation of prior distributions for Bayesian models and imputation of missing data. We introduce a framework that leverages LLMs to enhance Bayesian workflows by eliciting expert-like prior knowledge and imputing missing data. Tested on diverse datasets, this approach can improve predictive accuracy and reduce data requirements, offering significant potential in healthcare, environmental science and engineering applications. We discuss the implications and challenges of treating LLMs as 'experts'.
format Preprint
id arxiv_https___arxiv_org_abs_2402_07770
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Had enough of experts? Quantitative knowledge retrieval from large language models
Selby, David
Spriestersbach, Kai
Iwashita, Yuichiro
Saad, Mohammad
Bappert, Dennis
Warrier, Archana
Mukherjee, Sumantrak
Kise, Koichi
Vollmer, Sebastian
Information Retrieval
Computation and Language
Applications
Large language models (LLMs) have been extensively studied for their abilities to generate convincing natural language sequences, however their utility for quantitative information retrieval is less well understood. Here we explore the feasibility of LLMs as a mechanism for quantitative knowledge retrieval to aid two data analysis tasks: elicitation of prior distributions for Bayesian models and imputation of missing data. We introduce a framework that leverages LLMs to enhance Bayesian workflows by eliciting expert-like prior knowledge and imputing missing data. Tested on diverse datasets, this approach can improve predictive accuracy and reduce data requirements, offering significant potential in healthcare, environmental science and engineering applications. We discuss the implications and challenges of treating LLMs as 'experts'.
title Had enough of experts? Quantitative knowledge retrieval from large language models
topic Information Retrieval
Computation and Language
Applications
url https://arxiv.org/abs/2402.07770