@inproceedings{mahendra-etal-2025-evaluating,
title = "Evaluating Numeracy of Language Models as a Natural Language Inference Task",
author = "Mahendra, Rahmad and
Spina, Damiano and
Cavedon, Lawrence and
Verspoor, Karin",
editor = "Chiruzzo, Luis and
Ritter, Alan and
Wang, Lu",
booktitle = "Findings of the Association for Computational Linguistics: NAACL 2025",
month = apr,
year = "2025",
address = "Albuquerque, New Mexico",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.findings-naacl.467/",
doi = "10.18653/v1/2025.findings-naacl.467",
pages = "8336--8361",
ISBN = "979-8-89176-195-7",
abstract = "While recent advancements in large language models (LLMs) have enhanced their capabilities to solve mathematical problems, other aspects of numeracy remain underexplored. In this paper, we propose a benchmark to evaluate the ability of language models to perform basic numeracy tasks. We frame numeracy as a Natural Language Inference (NLI) task to assess the models' ability to understand both numbers and language contexts. We evaluate 49 language models (LMs), including fine-tuned LMs on NLI datasets, instruction-tuned LLMs, and specialized math-LLMs. Our findings reveal three main insights: (1) LLMs only clearly outperform smaller LMs in arithmetic tasks, indicating that mathematical reasoning cannot be generalized to other numeracy skills such as number comparison and normalization; (2) while most language models achieve fair to good accuracy for NLI entailment cases, they still struggle to predict contradiction and neutral cases; and (3) the robustness of language models' numeracy capabilities needs improvement, particularly in understanding the semantics and pragmatics of numbers in linguistic contexts."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="mahendra-etal-2025-evaluating">
<titleInfo>
<title>Evaluating Numeracy of Language Models as a Natural Language Inference Task</title>
</titleInfo>
<name type="personal">
<namePart type="given">Rahmad</namePart>
<namePart type="family">Mahendra</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Damiano</namePart>
<namePart type="family">Spina</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lawrence</namePart>
<namePart type="family">Cavedon</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Karin</namePart>
<namePart type="family">Verspoor</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-04</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: NAACL 2025</title>
</titleInfo>
<name type="personal">
<namePart type="given">Luis</namePart>
<namePart type="family">Chiruzzo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alan</namePart>
<namePart type="family">Ritter</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lu</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Albuquerque, New Mexico</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-195-7</identifier>
</relatedItem>
<abstract>While recent advancements in large language models (LLMs) have enhanced their capabilities to solve mathematical problems, other aspects of numeracy remain underexplored. In this paper, we propose a benchmark to evaluate the ability of language models to perform basic numeracy tasks. We frame numeracy as a Natural Language Inference (NLI) task to assess the models’ ability to understand both numbers and language contexts. We evaluate 49 language models (LMs), including fine-tuned LMs on NLI datasets, instruction-tuned LLMs, and specialized math-LLMs. Our findings reveal three main insights: (1) LLMs only clearly outperform smaller LMs in arithmetic tasks, indicating that mathematical reasoning cannot be generalized to other numeracy skills such as number comparison and normalization; (2) while most language models achieve fair to good accuracy for NLI entailment cases, they still struggle to predict contradiction and neutral cases; and (3) the robustness of language models’ numeracy capabilities needs improvement, particularly in understanding the semantics and pragmatics of numbers in linguistic contexts.</abstract>
<identifier type="citekey">mahendra-etal-2025-evaluating</identifier>
<identifier type="doi">10.18653/v1/2025.findings-naacl.467</identifier>
<location>
<url>https://aclanthology.org/2025.findings-naacl.467/</url>
</location>
<part>
<date>2025-04</date>
<extent unit="page">
<start>8336</start>
<end>8361</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Evaluating Numeracy of Language Models as a Natural Language Inference Task
%A Mahendra, Rahmad
%A Spina, Damiano
%A Cavedon, Lawrence
%A Verspoor, Karin
%Y Chiruzzo, Luis
%Y Ritter, Alan
%Y Wang, Lu
%S Findings of the Association for Computational Linguistics: NAACL 2025
%D 2025
%8 April
%I Association for Computational Linguistics
%C Albuquerque, New Mexico
%@ 979-8-89176-195-7
%F mahendra-etal-2025-evaluating
%X While recent advancements in large language models (LLMs) have enhanced their capabilities to solve mathematical problems, other aspects of numeracy remain underexplored. In this paper, we propose a benchmark to evaluate the ability of language models to perform basic numeracy tasks. We frame numeracy as a Natural Language Inference (NLI) task to assess the models’ ability to understand both numbers and language contexts. We evaluate 49 language models (LMs), including fine-tuned LMs on NLI datasets, instruction-tuned LLMs, and specialized math-LLMs. Our findings reveal three main insights: (1) LLMs only clearly outperform smaller LMs in arithmetic tasks, indicating that mathematical reasoning cannot be generalized to other numeracy skills such as number comparison and normalization; (2) while most language models achieve fair to good accuracy for NLI entailment cases, they still struggle to predict contradiction and neutral cases; and (3) the robustness of language models’ numeracy capabilities needs improvement, particularly in understanding the semantics and pragmatics of numbers in linguistic contexts.
%R 10.18653/v1/2025.findings-naacl.467
%U https://aclanthology.org/2025.findings-naacl.467/
%U https://doi.org/10.18653/v1/2025.findings-naacl.467
%P 8336-8361
Markdown (Informal)
[Evaluating Numeracy of Language Models as a Natural Language Inference Task](https://aclanthology.org/2025.findings-naacl.467/) (Mahendra et al., Findings 2025)
ACL