@inproceedings{nacar-etal-2026-gate-general,
title = "{GATE}: General {A}rabic Text Embedding for Enhanced Semantic Textual Similarity with Matryoshka Representation Learning and Hybrid Loss Training",
author = "Nacar, Omer and
Koubaa, Anis and
Sibaee, Serry and
Alhabashi, Yasser and
Ammar, Adel and
Boulila, Wadii",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.1952/",
pages = "41623--41634",
ISBN = "979-8-89176-462-0",
abstract = "Semantic textual similarity (STS) is central to retrieval, clustering, and semantic understanding, yet Arabic remains underserved by high-quality sentence-embedding models and evaluation resources. We introduce \textbf{GATE}, a compact family of Arabic sentence embedding models and a controlled study of Arabic-specific backbones, contrastive Matryoshka Representation Learning, dimensional truncation, and hybrid STS-NLI supervision. We evaluate the resulting models across Arabic semantic similarity, NLI, pair classification, sentence classification, clustering, retrieval, reranking, and dialectal semantic relatedness, against strong multilingual baselines including BGE-M3, Arctic-Embed-v2, and Qwen3-Embedding. Despite their 135M-parameter scale, GATE-family models are competitive with substantially larger multilingual encoders and achieve strong results on several native-Arabic semantic tasks. A three-annotator audit further finds high Arabic surface adequacy and 84.2{\%} majority relation preservation for the translated supervision. The released checkpoints, data resources, and experimental artifacts support reproducible Arabic embedding research."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="nacar-etal-2026-gate-general">
<titleInfo>
<title>GATE: General Arabic Text Embedding for Enhanced Semantic Textual Similarity with Matryoshka Representation Learning and Hybrid Loss Training</title>
</titleInfo>
<name type="personal">
<namePart type="given">Omer</namePart>
<namePart type="family">Nacar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Anis</namePart>
<namePart type="family">Koubaa</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Serry</namePart>
<namePart type="family">Sibaee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yasser</namePart>
<namePart type="family">Alhabashi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Adel</namePart>
<namePart type="family">Ammar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wadii</namePart>
<namePart type="family">Boulila</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>Semantic textual similarity (STS) is central to retrieval, clustering, and semantic understanding, yet Arabic remains underserved by high-quality sentence-embedding models and evaluation resources. We introduce GATE, a compact family of Arabic sentence embedding models and a controlled study of Arabic-specific backbones, contrastive Matryoshka Representation Learning, dimensional truncation, and hybrid STS-NLI supervision. We evaluate the resulting models across Arabic semantic similarity, NLI, pair classification, sentence classification, clustering, retrieval, reranking, and dialectal semantic relatedness, against strong multilingual baselines including BGE-M3, Arctic-Embed-v2, and Qwen3-Embedding. Despite their 135M-parameter scale, GATE-family models are competitive with substantially larger multilingual encoders and achieve strong results on several native-Arabic semantic tasks. A three-annotator audit further finds high Arabic surface adequacy and 84.2% majority relation preservation for the translated supervision. The released checkpoints, data resources, and experimental artifacts support reproducible Arabic embedding research.</abstract>
<identifier type="citekey">nacar-etal-2026-gate-general</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.1952/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>41623</start>
<end>41634</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T GATE: General Arabic Text Embedding for Enhanced Semantic Textual Similarity with Matryoshka Representation Learning and Hybrid Loss Training
%A Nacar, Omer
%A Koubaa, Anis
%A Sibaee, Serry
%A Alhabashi, Yasser
%A Ammar, Adel
%A Boulila, Wadii
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F nacar-etal-2026-gate-general
%X Semantic textual similarity (STS) is central to retrieval, clustering, and semantic understanding, yet Arabic remains underserved by high-quality sentence-embedding models and evaluation resources. We introduce GATE, a compact family of Arabic sentence embedding models and a controlled study of Arabic-specific backbones, contrastive Matryoshka Representation Learning, dimensional truncation, and hybrid STS-NLI supervision. We evaluate the resulting models across Arabic semantic similarity, NLI, pair classification, sentence classification, clustering, retrieval, reranking, and dialectal semantic relatedness, against strong multilingual baselines including BGE-M3, Arctic-Embed-v2, and Qwen3-Embedding. Despite their 135M-parameter scale, GATE-family models are competitive with substantially larger multilingual encoders and achieve strong results on several native-Arabic semantic tasks. A three-annotator audit further finds high Arabic surface adequacy and 84.2% majority relation preservation for the translated supervision. The released checkpoints, data resources, and experimental artifacts support reproducible Arabic embedding research.
%U https://aclanthology.org/2026.findings-emnlp.1952/
%P 41623-41634
Markdown (Informal)
[GATE: General Arabic Text Embedding for Enhanced Semantic Textual Similarity with Matryoshka Representation Learning and Hybrid Loss Training](https://aclanthology.org/2026.findings-emnlp.1952/) (Nacar et al., Findings 2026)
ACL