@inproceedings{arici-etal-2025-learning,
title = "Learning Trajectories of Figurative Language for Pre-Trained Language Models",
author = "Arici, Nicola and
Putelli, Luca and
Gjinika, Ejdis and
Serina, Ivan and
Gerevini, Alfonso Emilio",
editor = "Christodoulopoulos, Christos and
Chakraborty, Tanmoy and
Rose, Carolyn and
Peng, Violet",
booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2025",
month = nov,
year = "2025",
address = "Suzhou, China",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.findings-emnlp.779/",
doi = "10.18653/v1/2025.findings-emnlp.779",
pages = "14440--14461",
ISBN = "979-8-89176-335-7",
abstract = "Figurative language and figures of speech, such as metaphors and hyperboles, are used every day in written and oral communication among human beings. Nonetheless, this imaginative use of words in a non literal way requires a solid understanding of semantics and a deep real-world knowledge. In the longstanding debate about whether Neural Language Models (NLMs) really have a full understanding of text, analysing how they can recognise figurative language can provide some intuition of their functioning, their capabilities and their limits. Therefore, in this paper, we exploit probing tasks to study how several NLMs of different sizes recognise four different figures of speech: hyperboles, metaphors, oxymorons and pleonasms. We analyse whether this information is learned and how it is acquired during the training of the model, describing its learning trajectory. Moreover, we analyse which layers have a better comprehension of figurative language and the influence of pre-training data. Datasets and code are available at \url{https://github.com/nicolarici/learning-trajectories}."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="arici-etal-2025-learning">
<titleInfo>
<title>Learning Trajectories of Figurative Language for Pre-Trained Language Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Nicola</namePart>
<namePart type="family">Arici</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Luca</namePart>
<namePart type="family">Putelli</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ejdis</namePart>
<namePart type="family">Gjinika</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ivan</namePart>
<namePart type="family">Serina</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alfonso</namePart>
<namePart type="given">Emilio</namePart>
<namePart type="family">Gerevini</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2025</title>
</titleInfo>
<name type="personal">
<namePart type="given">Christos</namePart>
<namePart type="family">Christodoulopoulos</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tanmoy</namePart>
<namePart type="family">Chakraborty</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Carolyn</namePart>
<namePart type="family">Rose</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Violet</namePart>
<namePart type="family">Peng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Suzhou, China</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-335-7</identifier>
</relatedItem>
<abstract>Figurative language and figures of speech, such as metaphors and hyperboles, are used every day in written and oral communication among human beings. Nonetheless, this imaginative use of words in a non literal way requires a solid understanding of semantics and a deep real-world knowledge. In the longstanding debate about whether Neural Language Models (NLMs) really have a full understanding of text, analysing how they can recognise figurative language can provide some intuition of their functioning, their capabilities and their limits. Therefore, in this paper, we exploit probing tasks to study how several NLMs of different sizes recognise four different figures of speech: hyperboles, metaphors, oxymorons and pleonasms. We analyse whether this information is learned and how it is acquired during the training of the model, describing its learning trajectory. Moreover, we analyse which layers have a better comprehension of figurative language and the influence of pre-training data. Datasets and code are available at https://github.com/nicolarici/learning-trajectories.</abstract>
<identifier type="citekey">arici-etal-2025-learning</identifier>
<identifier type="doi">10.18653/v1/2025.findings-emnlp.779</identifier>
<location>
<url>https://aclanthology.org/2025.findings-emnlp.779/</url>
</location>
<part>
<date>2025-11</date>
<extent unit="page">
<start>14440</start>
<end>14461</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Learning Trajectories of Figurative Language for Pre-Trained Language Models
%A Arici, Nicola
%A Putelli, Luca
%A Gjinika, Ejdis
%A Serina, Ivan
%A Gerevini, Alfonso Emilio
%Y Christodoulopoulos, Christos
%Y Chakraborty, Tanmoy
%Y Rose, Carolyn
%Y Peng, Violet
%S Findings of the Association for Computational Linguistics: EMNLP 2025
%D 2025
%8 November
%I Association for Computational Linguistics
%C Suzhou, China
%@ 979-8-89176-335-7
%F arici-etal-2025-learning
%X Figurative language and figures of speech, such as metaphors and hyperboles, are used every day in written and oral communication among human beings. Nonetheless, this imaginative use of words in a non literal way requires a solid understanding of semantics and a deep real-world knowledge. In the longstanding debate about whether Neural Language Models (NLMs) really have a full understanding of text, analysing how they can recognise figurative language can provide some intuition of their functioning, their capabilities and their limits. Therefore, in this paper, we exploit probing tasks to study how several NLMs of different sizes recognise four different figures of speech: hyperboles, metaphors, oxymorons and pleonasms. We analyse whether this information is learned and how it is acquired during the training of the model, describing its learning trajectory. Moreover, we analyse which layers have a better comprehension of figurative language and the influence of pre-training data. Datasets and code are available at https://github.com/nicolarici/learning-trajectories.
%R 10.18653/v1/2025.findings-emnlp.779
%U https://aclanthology.org/2025.findings-emnlp.779/
%U https://doi.org/10.18653/v1/2025.findings-emnlp.779
%P 14440-14461
Markdown (Informal)
[Learning Trajectories of Figurative Language for Pre-Trained Language Models](https://aclanthology.org/2025.findings-emnlp.779/) (Arici et al., Findings 2025)
ACL