@inproceedings{park-kim-2025-llms,
title = "Where do {LLM}s Encode the Knowledge to Assess the Ambiguity?",
author = "Park, Hancheol and
Kim, Geonmin",
editor = "Rambow, Owen and
Wanner, Leo and
Apidianaki, Marianna and
Al-Khalifa, Hend and
Eugenio, Barbara Di and
Schockaert, Steven and
Darwish, Kareem and
Agarwal, Apoorv",
booktitle = "Proceedings of the 31st International Conference on Computational Linguistics: Industry Track",
month = jan,
year = "2025",
address = "Abu Dhabi, UAE",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.coling-industry.38/",
pages = "445--452",
abstract = "Recently, large language models (LLMs) have shown remarkable performance across various natural language processing tasks, thanks to their vast amount of knowledge. Nevertheless, they often generate unreliable responses. A common example is providing a single biased answer to an ambiguous question that could have multiple correct answers. To address this issue, in this study, we discuss methods to detect such ambiguous samples. More specifically, we propose a classifier that uses a representation from an intermediate layer of the LLM as input. This is based on observations from previous research that representations of ambiguous samples in intermediate layers are closer to those of relevant label samples in the embedding space, but not necessarily in higher layers. The experimental results demonstrate that using representations from intermediate layers detects ambiguous input prompts more effectively than using representations from the final layer. Furthermore, in this study, we propose a method to train such classifiers without ambiguity labels, as most datasets lack labels regarding the ambiguity of samples, and evaluate its effectiveness."
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="park-kim-2025-llms">
<titleInfo>
<title>Where do LLMs Encode the Knowledge to Assess the Ambiguity?</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hancheol</namePart>
<namePart type="family">Park</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Geonmin</namePart>
<namePart type="family">Kim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-01</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 31st International Conference on Computational Linguistics: Industry Track</title>
</titleInfo>
<name type="personal">
<namePart type="given">Owen</namePart>
<namePart type="family">Rambow</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Leo</namePart>
<namePart type="family">Wanner</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Marianna</namePart>
<namePart type="family">Apidianaki</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hend</namePart>
<namePart type="family">Al-Khalifa</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Barbara</namePart>
<namePart type="given">Di</namePart>
<namePart type="family">Eugenio</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Steven</namePart>
<namePart type="family">Schockaert</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kareem</namePart>
<namePart type="family">Darwish</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Apoorv</namePart>
<namePart type="family">Agarwal</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Abu Dhabi, UAE</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Recently, large language models (LLMs) have shown remarkable performance across various natural language processing tasks, thanks to their vast amount of knowledge. Nevertheless, they often generate unreliable responses. A common example is providing a single biased answer to an ambiguous question that could have multiple correct answers. To address this issue, in this study, we discuss methods to detect such ambiguous samples. More specifically, we propose a classifier that uses a representation from an intermediate layer of the LLM as input. This is based on observations from previous research that representations of ambiguous samples in intermediate layers are closer to those of relevant label samples in the embedding space, but not necessarily in higher layers. The experimental results demonstrate that using representations from intermediate layers detects ambiguous input prompts more effectively than using representations from the final layer. Furthermore, in this study, we propose a method to train such classifiers without ambiguity labels, as most datasets lack labels regarding the ambiguity of samples, and evaluate its effectiveness.</abstract>
<identifier type="citekey">park-kim-2025-llms</identifier>
<location>
<url>https://aclanthology.org/2025.coling-industry.38/</url>
</location>
<part>
<date>2025-01</date>
<extent unit="page">
<start>445</start>
<end>452</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Where do LLMs Encode the Knowledge to Assess the Ambiguity?
%A Park, Hancheol
%A Kim, Geonmin
%Y Rambow, Owen
%Y Wanner, Leo
%Y Apidianaki, Marianna
%Y Al-Khalifa, Hend
%Y Eugenio, Barbara Di
%Y Schockaert, Steven
%Y Darwish, Kareem
%Y Agarwal, Apoorv
%S Proceedings of the 31st International Conference on Computational Linguistics: Industry Track
%D 2025
%8 January
%I Association for Computational Linguistics
%C Abu Dhabi, UAE
%F park-kim-2025-llms
%X Recently, large language models (LLMs) have shown remarkable performance across various natural language processing tasks, thanks to their vast amount of knowledge. Nevertheless, they often generate unreliable responses. A common example is providing a single biased answer to an ambiguous question that could have multiple correct answers. To address this issue, in this study, we discuss methods to detect such ambiguous samples. More specifically, we propose a classifier that uses a representation from an intermediate layer of the LLM as input. This is based on observations from previous research that representations of ambiguous samples in intermediate layers are closer to those of relevant label samples in the embedding space, but not necessarily in higher layers. The experimental results demonstrate that using representations from intermediate layers detects ambiguous input prompts more effectively than using representations from the final layer. Furthermore, in this study, we propose a method to train such classifiers without ambiguity labels, as most datasets lack labels regarding the ambiguity of samples, and evaluate its effectiveness.
%U https://aclanthology.org/2025.coling-industry.38/
%P 445-452
Markdown (Informal)
[Where do LLMs Encode the Knowledge to Assess the Ambiguity?](https://aclanthology.org/2025.coling-industry.38/) (Park & Kim, COLING 2025)
ACL