<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="panchal-etal-2026-indic">
    <titleInfo>
        <title>Indic-TunedLens: Interpreting Multilingual Models in Indian Languages</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Mihir</namePart>
        <namePart type="family">Panchal</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Deeksha</namePart>
        <namePart type="family">Varshney</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Mamta</namePart>
        <namePart type="family">.</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Asif</namePart>
        <namePart type="family">Ekbal</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2026-03</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the 13th Workshop on NLP for Similar Languages, Varieties and Dialects</title>
        </titleInfo>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Rabat, Morocco</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
    </relatedItem>
    <abstract>Multilingual large language models (LLMs) are increasingly deployed in linguistically diverse regions like India, yet most interpretability tools remain tailored to English. Prior work reveals that LLMs often operate in English centric representation spaces, making cross lingual interpretability a pressing concern. We introduce Indic-TunedLens, a novel interpretability framework specifically for Indian languages that learns shared affine transformations. Unlike the standard Logit Lens, which directly decodes intermediate activations, Indic-TunedLens adjusts hidden states for each target language, aligning them with the target output distributions to enable more faithful decoding of model representations. We evaluate our framework on 10 Indian languages using the MMLU benchmark and find that it significantly improves over SOTA interpretability methods, especially for morphologically rich, low resource languages. Our results provide crucial insights into the layer-wise semantic encoding of multilingual transformers. Our model is available at https://huggingface.co/spaces/MihirRajeshPanchal/IndicTunedLens. Our code is available at https://github.com/MihirRajeshPanchal/IndicTunedLens.</abstract>
    <identifier type="citekey">panchal-etal-2026-indic</identifier>
    <location>
        <url>https://aclanthology.org/2026.vardial-1.14/</url>
    </location>
    <part>
        <date>2026-03</date>
        <extent unit="page">
            <start>172</start>
            <end>185</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Indic-TunedLens: Interpreting Multilingual Models in Indian Languages
%A Panchal, Mihir
%A Varshney, Deeksha
%A ., Mamta
%A Ekbal, Asif
%S Proceedings of the 13th Workshop on NLP for Similar Languages, Varieties and Dialects
%D 2026
%8 March
%I Association for Computational Linguistics
%C Rabat, Morocco
%F panchal-etal-2026-indic
%X Multilingual large language models (LLMs) are increasingly deployed in linguistically diverse regions like India, yet most interpretability tools remain tailored to English. Prior work reveals that LLMs often operate in English centric representation spaces, making cross lingual interpretability a pressing concern. We introduce Indic-TunedLens, a novel interpretability framework specifically for Indian languages that learns shared affine transformations. Unlike the standard Logit Lens, which directly decodes intermediate activations, Indic-TunedLens adjusts hidden states for each target language, aligning them with the target output distributions to enable more faithful decoding of model representations. We evaluate our framework on 10 Indian languages using the MMLU benchmark and find that it significantly improves over SOTA interpretability methods, especially for morphologically rich, low resource languages. Our results provide crucial insights into the layer-wise semantic encoding of multilingual transformers. Our model is available at https://huggingface.co/spaces/MihirRajeshPanchal/IndicTunedLens. Our code is available at https://github.com/MihirRajeshPanchal/IndicTunedLens.
%U https://aclanthology.org/2026.vardial-1.14/
%P 172-185

Markdown (Informal)

[Indic-TunedLens: Interpreting Multilingual Models in Indian Languages](https://aclanthology.org/2026.vardial-1.14/) (Panchal et al., VarDial 2026)

• Indic-TunedLens: Interpreting Multilingual Models in Indian Languages (Panchal et al., VarDial 2026)

ACL

• Mihir Panchal, Deeksha Varshney, Mamta ., and Asif Ekbal. 2026. Indic-TunedLens: Interpreting Multilingual Models in Indian Languages. In Proceedings of the 13th Workshop on NLP for Similar Languages, Varieties and Dialects, pages 172–185, Rabat, Morocco. Association for Computational Linguistics.