@inproceedings{kim-etal-2026-clulab,
title = "clulab-retrieval at {S}em{E}val-2026 Task 8: A Comparative Analysis of Dense Retrievers and {H}y{DE} for Multi-Turn Conversational Retrieval",
author = "Kim, Hyungji and
Kondapaneni, Siva Rohit and
Bethard, Steven",
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-1.351/",
pages = "2787--2792",
ISBN = "979-8-89176-414-9",
abstract = "We present a comparative analysis of dense retrievers and retrieval strategies for multi-turn conversational retrieval in SemEval-2026 Task 8 (MTRAGEval). Our official submission employed a fine-tuned E5-based dense retriever (E5-FT, {\textasciitilde}110M parameters) with Hypothetical Document Embeddings (HyDE), achieving nDCG@5 of .3309, ranking 31 out of 38 systems. On the development set we also compared E5-FT versus BGE embeddings, dense-only versus hybrid retrieval strategies, and HyDE versus keyword extraction approaches. We found: (1) BGE (general-purpose, {\textasciitilde}110M) outperforms our domain-fine-tuned E5-FT ({\textasciitilde}110M) by 30.5{\%} on baseline retrieval, suggesting that model selection may matter more than domain-specific fine-tuning, (2) hybrid retrieval combining BM25 and dense methods provides complementary signals, with HyDE improving BM25 by 26.7{\%} and dense retrieval by 4.0{\%}, and (3) keyword-based query simplification degrades performance by 11-28{\%} across domains, validating HyDE{'}s approach of preserving semantic richness through passage-level text."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="kim-etal-2026-clulab">
<titleInfo>
<title>clulab-retrieval at SemEval-2026 Task 8: A Comparative Analysis of Dense Retrievers and HyDE for Multi-Turn Conversational Retrieval</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hyungji</namePart>
<namePart type="family">Kim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Siva</namePart>
<namePart type="given">Rohit</namePart>
<namePart type="family">Kondapaneni</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Steven</namePart>
<namePart type="family">Bethard</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>We present a comparative analysis of dense retrievers and retrieval strategies for multi-turn conversational retrieval in SemEval-2026 Task 8 (MTRAGEval). Our official submission employed a fine-tuned E5-based dense retriever (E5-FT, ~110M parameters) with Hypothetical Document Embeddings (HyDE), achieving nDCG@5 of .3309, ranking 31 out of 38 systems. On the development set we also compared E5-FT versus BGE embeddings, dense-only versus hybrid retrieval strategies, and HyDE versus keyword extraction approaches. We found: (1) BGE (general-purpose, ~110M) outperforms our domain-fine-tuned E5-FT (~110M) by 30.5% on baseline retrieval, suggesting that model selection may matter more than domain-specific fine-tuning, (2) hybrid retrieval combining BM25 and dense methods provides complementary signals, with HyDE improving BM25 by 26.7% and dense retrieval by 4.0%, and (3) keyword-based query simplification degrades performance by 11-28% across domains, validating HyDE’s approach of preserving semantic richness through passage-level text.</abstract>
<identifier type="citekey">kim-etal-2026-clulab</identifier>
<location>
<url>https://aclanthology.org/2026.semeval-1.351/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>2787</start>
<end>2792</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T clulab-retrieval at SemEval-2026 Task 8: A Comparative Analysis of Dense Retrievers and HyDE for Multi-Turn Conversational Retrieval
%A Kim, Hyungji
%A Kondapaneni, Siva Rohit
%A Bethard, Steven
%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 kim-etal-2026-clulab
%X We present a comparative analysis of dense retrievers and retrieval strategies for multi-turn conversational retrieval in SemEval-2026 Task 8 (MTRAGEval). Our official submission employed a fine-tuned E5-based dense retriever (E5-FT, ~110M parameters) with Hypothetical Document Embeddings (HyDE), achieving nDCG@5 of .3309, ranking 31 out of 38 systems. On the development set we also compared E5-FT versus BGE embeddings, dense-only versus hybrid retrieval strategies, and HyDE versus keyword extraction approaches. We found: (1) BGE (general-purpose, ~110M) outperforms our domain-fine-tuned E5-FT (~110M) by 30.5% on baseline retrieval, suggesting that model selection may matter more than domain-specific fine-tuning, (2) hybrid retrieval combining BM25 and dense methods provides complementary signals, with HyDE improving BM25 by 26.7% and dense retrieval by 4.0%, and (3) keyword-based query simplification degrades performance by 11-28% across domains, validating HyDE’s approach of preserving semantic richness through passage-level text.
%U https://aclanthology.org/2026.semeval-1.351/
%P 2787-2792
Markdown (Informal)
[clulab-retrieval at SemEval-2026 Task 8: A Comparative Analysis of Dense Retrievers and HyDE for Multi-Turn Conversational Retrieval](https://aclanthology.org/2026.semeval-1.351/) (Kim et al., SemEval 2026)
ACL