@inproceedings{avitan-etal-2025-practical,
title = "A Practical Method for Generating String Counterfactuals",
author = "Avitan, Matan and
Cotterell, Ryan and
Goldberg, Yoav and
Ravfogel, Shauli",
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.180/",
doi = "10.18653/v1/2025.findings-naacl.180",
pages = "3267--3286",
ISBN = "979-8-89176-195-7",
abstract = "Interventions targeting the representation space of language models (LMs) have emerged as an effective means to influence model behavior. Such methods are employed, for example, to eliminate or alter the encoding of demographic information such as gender within the model{'}s representations and, in so doing, create a counterfactual representation. However, because the intervention operates within the representation space, understanding precisely what aspects of the text it modifies poses a challenge. In this paper, we give a method to convert representation counterfactuals into string counterfactuals. We demonstrate that this approach enables us to analyze the linguistic alterations corresponding to a given representation space intervention and to interpret the features utilized to encode a specific concept. Moreover, the resulting counterfactuals can be used to mitigate bias in classification through data augmentation."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="avitan-etal-2025-practical">
<titleInfo>
<title>A Practical Method for Generating String Counterfactuals</title>
</titleInfo>
<name type="personal">
<namePart type="given">Matan</namePart>
<namePart type="family">Avitan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ryan</namePart>
<namePart type="family">Cotterell</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yoav</namePart>
<namePart type="family">Goldberg</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shauli</namePart>
<namePart type="family">Ravfogel</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>Interventions targeting the representation space of language models (LMs) have emerged as an effective means to influence model behavior. Such methods are employed, for example, to eliminate or alter the encoding of demographic information such as gender within the model’s representations and, in so doing, create a counterfactual representation. However, because the intervention operates within the representation space, understanding precisely what aspects of the text it modifies poses a challenge. In this paper, we give a method to convert representation counterfactuals into string counterfactuals. We demonstrate that this approach enables us to analyze the linguistic alterations corresponding to a given representation space intervention and to interpret the features utilized to encode a specific concept. Moreover, the resulting counterfactuals can be used to mitigate bias in classification through data augmentation.</abstract>
<identifier type="citekey">avitan-etal-2025-practical</identifier>
<identifier type="doi">10.18653/v1/2025.findings-naacl.180</identifier>
<location>
<url>https://aclanthology.org/2025.findings-naacl.180/</url>
</location>
<part>
<date>2025-04</date>
<extent unit="page">
<start>3267</start>
<end>3286</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T A Practical Method for Generating String Counterfactuals
%A Avitan, Matan
%A Cotterell, Ryan
%A Goldberg, Yoav
%A Ravfogel, Shauli
%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 avitan-etal-2025-practical
%X Interventions targeting the representation space of language models (LMs) have emerged as an effective means to influence model behavior. Such methods are employed, for example, to eliminate or alter the encoding of demographic information such as gender within the model’s representations and, in so doing, create a counterfactual representation. However, because the intervention operates within the representation space, understanding precisely what aspects of the text it modifies poses a challenge. In this paper, we give a method to convert representation counterfactuals into string counterfactuals. We demonstrate that this approach enables us to analyze the linguistic alterations corresponding to a given representation space intervention and to interpret the features utilized to encode a specific concept. Moreover, the resulting counterfactuals can be used to mitigate bias in classification through data augmentation.
%R 10.18653/v1/2025.findings-naacl.180
%U https://aclanthology.org/2025.findings-naacl.180/
%U https://doi.org/10.18653/v1/2025.findings-naacl.180
%P 3267-3286
Markdown (Informal)
[A Practical Method for Generating String Counterfactuals](https://aclanthology.org/2025.findings-naacl.180/) (Avitan et al., Findings 2025)
ACL
- Matan Avitan, Ryan Cotterell, Yoav Goldberg, and Shauli Ravfogel. 2025. A Practical Method for Generating String Counterfactuals. In Findings of the Association for Computational Linguistics: NAACL 2025, pages 3267–3286, Albuquerque, New Mexico. Association for Computational Linguistics.