@inproceedings{huanshuo-etal-2025-ctrla,
title = "{C}trl{A}: Adaptive Retrieval-Augmented Generation via Inherent Control",
author = "Huanshuo, Liu and
Zhang, Hao and
Guo, Zhijiang and
Wang, Jing and
Dong, Kuicai and
Li, Xiangyang and
Lee, Yi Quan and
Zhang, Cong and
Liu, Yong",
editor = "Che, Wanxiang and
Nabende, Joyce and
Shutova, Ekaterina and
Pilehvar, Mohammad Taher",
booktitle = "Findings of the Association for Computational Linguistics: ACL 2025",
month = jul,
year = "2025",
address = "Vienna, Austria",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.findings-acl.652/",
doi = "10.18653/v1/2025.findings-acl.652",
pages = "12592--12618",
ISBN = "979-8-89176-256-5",
abstract = "Retrieval-augmented generation (RAG) has emerged as a promising solution for mitigating hallucinations of large language models (LLMs) with retrieved external knowledge. Adaptive RAG enhances this approach by enabling dynamic retrieval during generation, activating retrieval only when the query exceeds LLM{'}s internal knowledge. Existing methods primarily focus on detecting LLM{'}s confidence via statistical uncertainty. Instead, we present the first attempts to solve adaptive RAG from a representation perspective and develop an inherent control-based framework, termed CtrlA. Specifically, we extract the features that represent the honesty and confidence directions of LLM and adopt them to control LLM behavior and guide retrieval timing decisions. We also design a simple yet effective query formulation strategy to support adaptive retrieval. Experiments show that CtrlA is superior to existing adaptive RAG methods on a diverse set of tasks. Honesty steering can effectively make LLMs more honest and confidence monitoring is a promising indicator of retrieval trigger."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="huanshuo-etal-2025-ctrla">
<titleInfo>
<title>CtrlA: Adaptive Retrieval-Augmented Generation via Inherent Control</title>
</titleInfo>
<name type="personal">
<namePart type="given">Liu</namePart>
<namePart type="family">Huanshuo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hao</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhijiang</namePart>
<namePart type="family">Guo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jing</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kuicai</namePart>
<namePart type="family">Dong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiangyang</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yi</namePart>
<namePart type="given">Quan</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Cong</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yong</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: ACL 2025</title>
</titleInfo>
<name type="personal">
<namePart type="given">Wanxiang</namePart>
<namePart type="family">Che</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Joyce</namePart>
<namePart type="family">Nabende</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ekaterina</namePart>
<namePart type="family">Shutova</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mohammad</namePart>
<namePart type="given">Taher</namePart>
<namePart type="family">Pilehvar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Vienna, Austria</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-256-5</identifier>
</relatedItem>
<abstract>Retrieval-augmented generation (RAG) has emerged as a promising solution for mitigating hallucinations of large language models (LLMs) with retrieved external knowledge. Adaptive RAG enhances this approach by enabling dynamic retrieval during generation, activating retrieval only when the query exceeds LLM’s internal knowledge. Existing methods primarily focus on detecting LLM’s confidence via statistical uncertainty. Instead, we present the first attempts to solve adaptive RAG from a representation perspective and develop an inherent control-based framework, termed CtrlA. Specifically, we extract the features that represent the honesty and confidence directions of LLM and adopt them to control LLM behavior and guide retrieval timing decisions. We also design a simple yet effective query formulation strategy to support adaptive retrieval. Experiments show that CtrlA is superior to existing adaptive RAG methods on a diverse set of tasks. Honesty steering can effectively make LLMs more honest and confidence monitoring is a promising indicator of retrieval trigger.</abstract>
<identifier type="citekey">huanshuo-etal-2025-ctrla</identifier>
<identifier type="doi">10.18653/v1/2025.findings-acl.652</identifier>
<location>
<url>https://aclanthology.org/2025.findings-acl.652/</url>
</location>
<part>
<date>2025-07</date>
<extent unit="page">
<start>12592</start>
<end>12618</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T CtrlA: Adaptive Retrieval-Augmented Generation via Inherent Control
%A Huanshuo, Liu
%A Zhang, Hao
%A Guo, Zhijiang
%A Wang, Jing
%A Dong, Kuicai
%A Li, Xiangyang
%A Lee, Yi Quan
%A Zhang, Cong
%A Liu, Yong
%Y Che, Wanxiang
%Y Nabende, Joyce
%Y Shutova, Ekaterina
%Y Pilehvar, Mohammad Taher
%S Findings of the Association for Computational Linguistics: ACL 2025
%D 2025
%8 July
%I Association for Computational Linguistics
%C Vienna, Austria
%@ 979-8-89176-256-5
%F huanshuo-etal-2025-ctrla
%X Retrieval-augmented generation (RAG) has emerged as a promising solution for mitigating hallucinations of large language models (LLMs) with retrieved external knowledge. Adaptive RAG enhances this approach by enabling dynamic retrieval during generation, activating retrieval only when the query exceeds LLM’s internal knowledge. Existing methods primarily focus on detecting LLM’s confidence via statistical uncertainty. Instead, we present the first attempts to solve adaptive RAG from a representation perspective and develop an inherent control-based framework, termed CtrlA. Specifically, we extract the features that represent the honesty and confidence directions of LLM and adopt them to control LLM behavior and guide retrieval timing decisions. We also design a simple yet effective query formulation strategy to support adaptive retrieval. Experiments show that CtrlA is superior to existing adaptive RAG methods on a diverse set of tasks. Honesty steering can effectively make LLMs more honest and confidence monitoring is a promising indicator of retrieval trigger.
%R 10.18653/v1/2025.findings-acl.652
%U https://aclanthology.org/2025.findings-acl.652/
%U https://doi.org/10.18653/v1/2025.findings-acl.652
%P 12592-12618
Markdown (Informal)
[CtrlA: Adaptive Retrieval-Augmented Generation via Inherent Control](https://aclanthology.org/2025.findings-acl.652/) (Huanshuo et al., Findings 2025)
ACL
- Liu Huanshuo, Hao Zhang, Zhijiang Guo, Jing Wang, Kuicai Dong, Xiangyang Li, Yi Quan Lee, Cong Zhang, and Yong Liu. 2025. CtrlA: Adaptive Retrieval-Augmented Generation via Inherent Control. In Findings of the Association for Computational Linguistics: ACL 2025, pages 12592–12618, Vienna, Austria. Association for Computational Linguistics.