@inproceedings{zhang-etal-2026-dont,
title = "Don{'}t Be Misled by Style: A Style-Adaptive Reranker for Capturing Effective Knowledge in Retrieval-Augmented Generation",
author = "Zhang, Ruwen and
Liu, Bo and
Xiang, Zhang Sheng and
Chen, Yida and
Zhao, Hantao and
Ding, Ding and
Jin, Jiahui and
Cao, Jiuxin",
editor = "Liakata, Maria and
Moreira, Viviane P. and
Zhang, Jiajun and
Jurgens, David",
booktitle = "Proceedings of the 64th Annual Meeting of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
month = jul,
year = "2026",
address = "San Diego, California, United States",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.acl-long.302/",
pages = "6665--6681",
ISBN = "979-8-89176-390-6",
abstract = "Rerankers are critical in Retrieval-Augmented Generation (RAG) for filtering evidence that enhances the accurate generation of LLMs. With the extension to open-domain scenarios, rerankers are inevitably deployed on mixed-style corpora, whereas most existing rerankers are mainly trained on well-edited texts. A rarely explored issue lies in enabling rerankers to maximally capture the effective knowledge for downstream LLMs without being misled by stylistic features. To address this issue, we propose SARK (Style-Adaptive Reranker with Knowledge Prioritization), a style-augmented multi-task framework that prioritizes effective knowledge over stylistic perturbations. SARK performs multi-granular knowledge mining by using an LLM to derive passage-level supervision on whether a passage helps or harms answer correctness, and list-level relative ranking preferences over candidate passages. It then jointly optimizes the reranker model with passage-level classification and list-level ranking objectives via style-augmented multi-task learning, encouraging the model to focus on the information needed for answering under mixed-style scenarios. Extensive experiments demonstrate that SARK improves generation performance across multiple LLMs under mixed-style conditions."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zhang-etal-2026-dont">
<titleInfo>
<title>Don’t Be Misled by Style: A Style-Adaptive Reranker for Capturing Effective Knowledge in Retrieval-Augmented Generation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ruwen</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Bo</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhang</namePart>
<namePart type="given">Sheng</namePart>
<namePart type="family">Xiang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yida</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hantao</namePart>
<namePart type="family">Zhao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ding</namePart>
<namePart type="family">Ding</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiahui</namePart>
<namePart type="family">Jin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiuxin</namePart>
<namePart type="family">Cao</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 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Maria</namePart>
<namePart type="family">Liakata</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Viviane</namePart>
<namePart type="given">P</namePart>
<namePart type="family">Moreira</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiajun</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">David</namePart>
<namePart type="family">Jurgens</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, United States</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-390-6</identifier>
</relatedItem>
<abstract>Rerankers are critical in Retrieval-Augmented Generation (RAG) for filtering evidence that enhances the accurate generation of LLMs. With the extension to open-domain scenarios, rerankers are inevitably deployed on mixed-style corpora, whereas most existing rerankers are mainly trained on well-edited texts. A rarely explored issue lies in enabling rerankers to maximally capture the effective knowledge for downstream LLMs without being misled by stylistic features. To address this issue, we propose SARK (Style-Adaptive Reranker with Knowledge Prioritization), a style-augmented multi-task framework that prioritizes effective knowledge over stylistic perturbations. SARK performs multi-granular knowledge mining by using an LLM to derive passage-level supervision on whether a passage helps or harms answer correctness, and list-level relative ranking preferences over candidate passages. It then jointly optimizes the reranker model with passage-level classification and list-level ranking objectives via style-augmented multi-task learning, encouraging the model to focus on the information needed for answering under mixed-style scenarios. Extensive experiments demonstrate that SARK improves generation performance across multiple LLMs under mixed-style conditions.</abstract>
<identifier type="citekey">zhang-etal-2026-dont</identifier>
<location>
<url>https://aclanthology.org/2026.acl-long.302/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>6665</start>
<end>6681</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Don’t Be Misled by Style: A Style-Adaptive Reranker for Capturing Effective Knowledge in Retrieval-Augmented Generation
%A Zhang, Ruwen
%A Liu, Bo
%A Xiang, Zhang Sheng
%A Chen, Yida
%A Zhao, Hantao
%A Ding, Ding
%A Jin, Jiahui
%A Cao, Jiuxin
%Y Liakata, Maria
%Y Moreira, Viviane P.
%Y Zhang, Jiajun
%Y Jurgens, David
%S Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, United States
%@ 979-8-89176-390-6
%F zhang-etal-2026-dont
%X Rerankers are critical in Retrieval-Augmented Generation (RAG) for filtering evidence that enhances the accurate generation of LLMs. With the extension to open-domain scenarios, rerankers are inevitably deployed on mixed-style corpora, whereas most existing rerankers are mainly trained on well-edited texts. A rarely explored issue lies in enabling rerankers to maximally capture the effective knowledge for downstream LLMs without being misled by stylistic features. To address this issue, we propose SARK (Style-Adaptive Reranker with Knowledge Prioritization), a style-augmented multi-task framework that prioritizes effective knowledge over stylistic perturbations. SARK performs multi-granular knowledge mining by using an LLM to derive passage-level supervision on whether a passage helps or harms answer correctness, and list-level relative ranking preferences over candidate passages. It then jointly optimizes the reranker model with passage-level classification and list-level ranking objectives via style-augmented multi-task learning, encouraging the model to focus on the information needed for answering under mixed-style scenarios. Extensive experiments demonstrate that SARK improves generation performance across multiple LLMs under mixed-style conditions.
%U https://aclanthology.org/2026.acl-long.302/
%P 6665-6681
Markdown (Informal)
[Don’t Be Misled by Style: A Style-Adaptive Reranker for Capturing Effective Knowledge in Retrieval-Augmented Generation](https://aclanthology.org/2026.acl-long.302/) (Zhang et al., ACL 2026)
ACL
- Ruwen Zhang, Bo Liu, Zhang Sheng Xiang, Yida Chen, Hantao Zhao, Ding Ding, Jiahui Jin, and Jiuxin Cao. 2026. Don’t Be Misled by Style: A Style-Adaptive Reranker for Capturing Effective Knowledge in Retrieval-Augmented Generation. In Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 6665–6681, San Diego, California, United States. Association for Computational Linguistics.