@inproceedings{ulli-kumari-2026-polar,
title = "{P}ol{AR} Bears at {S}em{E}val-2026 Task 9: Parameter-Efficient Fine-Tuning and Cross-Lingual Augmentation for Multilingual Polarization Detection",
author = "Ulli, Vinay and
Kumari, Jyoti",
editor = "Kochmar, Ekaterina and
Ghosh, Debanjan and
North, Kai and
Komachi, Mamoru",
booktitle = "Proceedings of the 20th {I}nternational {W}orkshop on {S}emantic {E}valuation (2026)",
month = jul,
year = "2026",
address = "San Diego, California, USA",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.semeval-1.279/",
pages = "2209--2215",
ISBN = "979-8-89176-414-9",
abstract = "This paper describes our system for SemEval-2026 Task 9: Detecting Multilingual, Multicul-tural and Multievent Online Polarization. Wefocus on four low-resource Indian languages(Hindi, Bengali, Telugu, and Odia) across threesubtasks: Polarization Detection, Type Classi-fication, and Manifestation Identification. Toaddress data scarcity, we employ cross-lingualdata augmentation using IndicTrans2, expand-ing our dataset fourfold. Our unified architec-ture leverages Qwen3-4B-Instruct optimizedvia QLoRA, training a linear classification headon masked mean-pooled hidden states withonly {\ensuremath{\sim}}33M trainable parameters. Our systemachieved highly competitive results in Subtask1, with an average Macro F1 of 0.813 across alllanguages (peaking at 0.8668 for Telugu). Forthe complex multi-label frameworks of Sub-tasks 2 and 3, our results expose a significantpre-training bias within foundational LLMs;while Hindi maintained strong F1 scores of0.7008 and 0.7248, performance dropped con-siderably for the other three languages, high-lighting the ongoing challenges of cross-lingualtransfer for nuanced rhetorical techniques."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="ulli-kumari-2026-polar">
<titleInfo>
<title>PolAR Bears at SemEval-2026 Task 9: Parameter-Efficient Fine-Tuning and Cross-Lingual Augmentation for Multilingual Polarization Detection</title>
</titleInfo>
<name type="personal">
<namePart type="given">Vinay</namePart>
<namePart type="family">Ulli</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jyoti</namePart>
<namePart type="family">Kumari</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 20th International Workshop on Semantic Evaluation (2026)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ekaterina</namePart>
<namePart type="family">Kochmar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Debanjan</namePart>
<namePart type="family">Ghosh</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kai</namePart>
<namePart type="family">North</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mamoru</namePart>
<namePart type="family">Komachi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">San Diego, California, USA</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-414-9</identifier>
</relatedItem>
<abstract>This paper describes our system for SemEval-2026 Task 9: Detecting Multilingual, Multicul-tural and Multievent Online Polarization. Wefocus on four low-resource Indian languages(Hindi, Bengali, Telugu, and Odia) across threesubtasks: Polarization Detection, Type Classi-fication, and Manifestation Identification. Toaddress data scarcity, we employ cross-lingualdata augmentation using IndicTrans2, expand-ing our dataset fourfold. Our unified architec-ture leverages Qwen3-4B-Instruct optimizedvia QLoRA, training a linear classification headon masked mean-pooled hidden states withonly \ensuremath\sim33M trainable parameters. Our systemachieved highly competitive results in Subtask1, with an average Macro F1 of 0.813 across alllanguages (peaking at 0.8668 for Telugu). Forthe complex multi-label frameworks of Sub-tasks 2 and 3, our results expose a significantpre-training bias within foundational LLMs;while Hindi maintained strong F1 scores of0.7008 and 0.7248, performance dropped con-siderably for the other three languages, high-lighting the ongoing challenges of cross-lingualtransfer for nuanced rhetorical techniques.</abstract>
<identifier type="citekey">ulli-kumari-2026-polar</identifier>
<location>
<url>https://aclanthology.org/2026.semeval-1.279/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>2209</start>
<end>2215</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T PolAR Bears at SemEval-2026 Task 9: Parameter-Efficient Fine-Tuning and Cross-Lingual Augmentation for Multilingual Polarization Detection
%A Ulli, Vinay
%A Kumari, Jyoti
%Y Kochmar, Ekaterina
%Y Ghosh, Debanjan
%Y North, Kai
%Y Komachi, Mamoru
%S Proceedings of the 20th International Workshop on Semantic Evaluation (2026)
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, USA
%@ 979-8-89176-414-9
%F ulli-kumari-2026-polar
%X This paper describes our system for SemEval-2026 Task 9: Detecting Multilingual, Multicul-tural and Multievent Online Polarization. Wefocus on four low-resource Indian languages(Hindi, Bengali, Telugu, and Odia) across threesubtasks: Polarization Detection, Type Classi-fication, and Manifestation Identification. Toaddress data scarcity, we employ cross-lingualdata augmentation using IndicTrans2, expand-ing our dataset fourfold. Our unified architec-ture leverages Qwen3-4B-Instruct optimizedvia QLoRA, training a linear classification headon masked mean-pooled hidden states withonly \ensuremath\sim33M trainable parameters. Our systemachieved highly competitive results in Subtask1, with an average Macro F1 of 0.813 across alllanguages (peaking at 0.8668 for Telugu). Forthe complex multi-label frameworks of Sub-tasks 2 and 3, our results expose a significantpre-training bias within foundational LLMs;while Hindi maintained strong F1 scores of0.7008 and 0.7248, performance dropped con-siderably for the other three languages, high-lighting the ongoing challenges of cross-lingualtransfer for nuanced rhetorical techniques.
%U https://aclanthology.org/2026.semeval-1.279/
%P 2209-2215
Markdown (Informal)
[PolAR Bears at SemEval-2026 Task 9: Parameter-Efficient Fine-Tuning and Cross-Lingual Augmentation for Multilingual Polarization Detection](https://aclanthology.org/2026.semeval-1.279/) (Ulli & Kumari, SemEval 2026)
ACL