@inproceedings{shibata-etal-2026-curiosai,
title = "{C}urios{AI} at {S}em{E}val-2026 Task 4: A Comprehensive Study of Zero-Shot versus Fine-Tuned Approaches for Narrative Similarity",
author = "Shibata, Yuki and
Takushima, Hiroki and
Beppu, Fumika and
Manoj Kumar, Aiswariya and
Yamaga, Daichi and
Hori, Takayuki",
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-1.62/",
pages = "434--439",
ISBN = "979-8-89176-414-9",
abstract = "This paper presents our system for SemEval-2026 Task 4 on narrative similarity assessment.Through comprehensive experimentation, we evaluated various approaches including zero-shot pre-trained models, prompt engineering with large language models, and multiple fine-tuning strategies using synthetic data. Our experiments revealed a surprising finding: pre-trained sentence transformers in a zero-shot setting consistently outperformed all fine-tuning attempts. Specifically, our best system using sentence-transformers/sentence-t5-xl achieved 67.5{\%} accuracy on the development set (95{\%} CI: [61.0{\%}, 74.0{\%}]), while all fine-tuning approaches resulted in performance degradation of 2-18 percentage points. We provide a detailed analysis of why fine-tuning failed and discuss the implications for narrative similarity tasks."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="shibata-etal-2026-curiosai">
<titleInfo>
<title>CuriosAI at SemEval-2026 Task 4: A Comprehensive Study of Zero-Shot versus Fine-Tuned Approaches for Narrative Similarity</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yuki</namePart>
<namePart type="family">Shibata</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hiroki</namePart>
<namePart type="family">Takushima</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Fumika</namePart>
<namePart type="family">Beppu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Aiswariya</namePart>
<namePart type="family">Manoj Kumar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Daichi</namePart>
<namePart type="family">Yamaga</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Takayuki</namePart>
<namePart type="family">Hori</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 presents our system for SemEval-2026 Task 4 on narrative similarity assessment.Through comprehensive experimentation, we evaluated various approaches including zero-shot pre-trained models, prompt engineering with large language models, and multiple fine-tuning strategies using synthetic data. Our experiments revealed a surprising finding: pre-trained sentence transformers in a zero-shot setting consistently outperformed all fine-tuning attempts. Specifically, our best system using sentence-transformers/sentence-t5-xl achieved 67.5% accuracy on the development set (95% CI: [61.0%, 74.0%]), while all fine-tuning approaches resulted in performance degradation of 2-18 percentage points. We provide a detailed analysis of why fine-tuning failed and discuss the implications for narrative similarity tasks.</abstract>
<identifier type="citekey">shibata-etal-2026-curiosai</identifier>
<location>
<url>https://aclanthology.org/2026.semeval-1.62/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>434</start>
<end>439</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T CuriosAI at SemEval-2026 Task 4: A Comprehensive Study of Zero-Shot versus Fine-Tuned Approaches for Narrative Similarity
%A Shibata, Yuki
%A Takushima, Hiroki
%A Beppu, Fumika
%A Manoj Kumar, Aiswariya
%A Yamaga, Daichi
%A Hori, Takayuki
%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 shibata-etal-2026-curiosai
%X This paper presents our system for SemEval-2026 Task 4 on narrative similarity assessment.Through comprehensive experimentation, we evaluated various approaches including zero-shot pre-trained models, prompt engineering with large language models, and multiple fine-tuning strategies using synthetic data. Our experiments revealed a surprising finding: pre-trained sentence transformers in a zero-shot setting consistently outperformed all fine-tuning attempts. Specifically, our best system using sentence-transformers/sentence-t5-xl achieved 67.5% accuracy on the development set (95% CI: [61.0%, 74.0%]), while all fine-tuning approaches resulted in performance degradation of 2-18 percentage points. We provide a detailed analysis of why fine-tuning failed and discuss the implications for narrative similarity tasks.
%U https://aclanthology.org/2026.semeval-1.62/
%P 434-439
Markdown (Informal)
[CuriosAI at SemEval-2026 Task 4: A Comprehensive Study of Zero-Shot versus Fine-Tuned Approaches for Narrative Similarity](https://aclanthology.org/2026.semeval-1.62/) (Shibata et al., SemEval 2026)
ACL