Salvato in:
Dettagli Bibliografici
Autori principali: Keluskar, Aryan, Bhattacharjee, Amrita, Liu, Huan
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:https://arxiv.org/abs/2411.12395
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866916487878410240
author Keluskar, Aryan
Bhattacharjee, Amrita
Liu, Huan
author_facet Keluskar, Aryan
Bhattacharjee, Amrita
Liu, Huan
contents Ambiguity in natural language poses significant challenges to Large Language Models (LLMs) used for open-domain question answering. LLMs often struggle with the inherent uncertainties of human communication, leading to misinterpretations, miscommunications, hallucinations, and biased responses. This significantly weakens their ability to be used for tasks like fact-checking, question answering, feature extraction, and sentiment analysis. Using open-domain question answering as a test case, we compare off-the-shelf and few-shot LLM performance, focusing on measuring the impact of explicit disambiguation strategies. We demonstrate how simple, training-free, token-level disambiguation methods may be effectively used to improve LLM performance for ambiguous question answering tasks. We empirically show our findings and discuss best practices and broader impacts regarding ambiguity in LLMs.
format Preprint
id arxiv_https___arxiv_org_abs_2411_12395
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Do LLMs Understand Ambiguity in Text? A Case Study in Open-world Question Answering
Keluskar, Aryan
Bhattacharjee, Amrita
Liu, Huan
Computation and Language
Artificial Intelligence
Ambiguity in natural language poses significant challenges to Large Language Models (LLMs) used for open-domain question answering. LLMs often struggle with the inherent uncertainties of human communication, leading to misinterpretations, miscommunications, hallucinations, and biased responses. This significantly weakens their ability to be used for tasks like fact-checking, question answering, feature extraction, and sentiment analysis. Using open-domain question answering as a test case, we compare off-the-shelf and few-shot LLM performance, focusing on measuring the impact of explicit disambiguation strategies. We demonstrate how simple, training-free, token-level disambiguation methods may be effectively used to improve LLM performance for ambiguous question answering tasks. We empirically show our findings and discuss best practices and broader impacts regarding ambiguity in LLMs.
title Do LLMs Understand Ambiguity in Text? A Case Study in Open-world Question Answering
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2411.12395