@inproceedings{revankar-etal-2026-slugrag,
title = "{S}lug{RAG} at {S}em{E}val-2026 Task 8: Domain-Specific Fine-Tuning and Model Scaling for Multi-Turn {RAG} Retrieval",
author = "Revankar, Pratibha and
Kim, Jihye and
Azirakhmet, Umit",
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-1.135/",
pages = "981--987",
ISBN = "979-8-89176-414-9",
abstract = "Multi-Turn Retrieval-Augmented Generation (MT-RAG) requires resolving context-dependent ambiguities across conversational turns. We present a systematic evaluation of dense retrieval optimization for the MTRAGEval benchmark (Task 8, Subtask A: Retrieval Only), investigating training-time strategies and inference-time query reformulation across four diverse English-language domains: CLAPNQ (legal/patent), FIQA (financial), GOVT (government documents), and CLOUD (cloud computing). Our experiments demonstrate that domain-specific fine-tuning yields the most substantial gains, with our best CLAPNQ model achieving Recall@10 of 0.6016 and nDCG@10 of 0.4981{---}representing 58.3{\textbackslash}{\%} and 66.0{\textbackslash}{\%} improvements over the pre-trained BGE baseline. Domain-specific models average 44.3{\textbackslash}{\%} improvement in Recall@10 and 47.8{\textbackslash}{\%} in nDCG@10 across all domains. Additionally, fine-tuning larger embedding models (BGE-large) achieves the best overall performance (nDCG@10: 0.5101, Recall@10: 0.6221), highlighting the complementary impact of model capacity and domain adaptation."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="revankar-etal-2026-slugrag">
<titleInfo>
<title>SlugRAG at SemEval-2026 Task 8: Domain-Specific Fine-Tuning and Model Scaling for Multi-Turn RAG Retrieval</title>
</titleInfo>
<name type="personal">
<namePart type="given">Pratibha</namePart>
<namePart type="family">Revankar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jihye</namePart>
<namePart type="family">Kim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Umit</namePart>
<namePart type="family">Azirakhmet</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>Multi-Turn Retrieval-Augmented Generation (MT-RAG) requires resolving context-dependent ambiguities across conversational turns. We present a systematic evaluation of dense retrieval optimization for the MTRAGEval benchmark (Task 8, Subtask A: Retrieval Only), investigating training-time strategies and inference-time query reformulation across four diverse English-language domains: CLAPNQ (legal/patent), FIQA (financial), GOVT (government documents), and CLOUD (cloud computing). Our experiments demonstrate that domain-specific fine-tuning yields the most substantial gains, with our best CLAPNQ model achieving Recall@10 of 0.6016 and nDCG@10 of 0.4981—representing 58.3\textbackslash% and 66.0\textbackslash% improvements over the pre-trained BGE baseline. Domain-specific models average 44.3\textbackslash% improvement in Recall@10 and 47.8\textbackslash% in nDCG@10 across all domains. Additionally, fine-tuning larger embedding models (BGE-large) achieves the best overall performance (nDCG@10: 0.5101, Recall@10: 0.6221), highlighting the complementary impact of model capacity and domain adaptation.</abstract>
<identifier type="citekey">revankar-etal-2026-slugrag</identifier>
<location>
<url>https://aclanthology.org/2026.semeval-1.135/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>981</start>
<end>987</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T SlugRAG at SemEval-2026 Task 8: Domain-Specific Fine-Tuning and Model Scaling for Multi-Turn RAG Retrieval
%A Revankar, Pratibha
%A Kim, Jihye
%A Azirakhmet, Umit
%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 revankar-etal-2026-slugrag
%X Multi-Turn Retrieval-Augmented Generation (MT-RAG) requires resolving context-dependent ambiguities across conversational turns. We present a systematic evaluation of dense retrieval optimization for the MTRAGEval benchmark (Task 8, Subtask A: Retrieval Only), investigating training-time strategies and inference-time query reformulation across four diverse English-language domains: CLAPNQ (legal/patent), FIQA (financial), GOVT (government documents), and CLOUD (cloud computing). Our experiments demonstrate that domain-specific fine-tuning yields the most substantial gains, with our best CLAPNQ model achieving Recall@10 of 0.6016 and nDCG@10 of 0.4981—representing 58.3\textbackslash% and 66.0\textbackslash% improvements over the pre-trained BGE baseline. Domain-specific models average 44.3\textbackslash% improvement in Recall@10 and 47.8\textbackslash% in nDCG@10 across all domains. Additionally, fine-tuning larger embedding models (BGE-large) achieves the best overall performance (nDCG@10: 0.5101, Recall@10: 0.6221), highlighting the complementary impact of model capacity and domain adaptation.
%U https://aclanthology.org/2026.semeval-1.135/
%P 981-987
Markdown (Informal)
[SlugRAG at SemEval-2026 Task 8: Domain-Specific Fine-Tuning and Model Scaling for Multi-Turn RAG Retrieval](https://aclanthology.org/2026.semeval-1.135/) (Revankar et al., SemEval 2026)
ACL