@inproceedings{han-etal-2026-errv,
title = "{ERRV}: Eliciting Efficient Reasoning through Reasoning Vectors for Policy Optimization in Large Language Models",
author = "Han, Zhuowen and
Yang, Lei and
Jin, Renren and
Shi, Dan and
Sun, Chenxi and
Xiong, Deyi",
editor = "Liakata, Maria and
Moreira, Viviane P. and
Zhang, Jiajun and
Jurgens, David",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {ACL} 2026",
month = jul,
year = "2026",
address = "San Diego, California, United States",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-acl.1425/",
pages = "28557--28570",
ISBN = "979-8-89176-395-1",
abstract = "Recently, large reasoning models have achieved impressive performance, but their lengthy reasoning processes incur substantial inference overhead. To mitigate this issue, we propose the concept of reasoning vectors, representations extracted from the model{'}s hidden states, which can guide the model towards generating more concise and accurate responses. Building upon this, we present ERRV, a training framework that elicits efficient reasoning through reasoning vectors, which enables the model to generate high-quality responses during reinforcement learning. By performing targeted policy optimization on both accuracy and length objectives, ERRV effectively activates the model{'}s latent capability for efficient reasoning. Our experiments demonstrate that after training with ERRV, the model achieves approximately 30{\%} reduction in reasoning length while maintaining stable accuracy, without guidance from the reasoning vector during inference. This establishes a trade-off between efficiency and performance. Furthermore, we identify key properties of reasoning vectors: robustness, characterized by high similarity before and after training, and generalizability, demonstrating applicability across base models, distilled models, RL-trained models, parameter-merged models, and mixed-thought models. These properties collectively guarantee the reliability and broad applicability of our approach."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="han-etal-2026-errv">
<titleInfo>
<title>ERRV: Eliciting Efficient Reasoning through Reasoning Vectors for Policy Optimization in Large Language Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Zhuowen</namePart>
<namePart type="family">Han</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lei</namePart>
<namePart type="family">Yang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Renren</namePart>
<namePart type="family">Jin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dan</namePart>
<namePart type="family">Shi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chenxi</namePart>
<namePart type="family">Sun</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Deyi</namePart>
<namePart type="family">Xiong</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>Findings of the Association for Computational Linguistics: ACL 2026</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">979-8-89176-395-1</identifier>
</relatedItem>
<abstract>Recently, large reasoning models have achieved impressive performance, but their lengthy reasoning processes incur substantial inference overhead. To mitigate this issue, we propose the concept of reasoning vectors, representations extracted from the model’s hidden states, which can guide the model towards generating more concise and accurate responses. Building upon this, we present ERRV, a training framework that elicits efficient reasoning through reasoning vectors, which enables the model to generate high-quality responses during reinforcement learning. By performing targeted policy optimization on both accuracy and length objectives, ERRV effectively activates the model’s latent capability for efficient reasoning. Our experiments demonstrate that after training with ERRV, the model achieves approximately 30% reduction in reasoning length while maintaining stable accuracy, without guidance from the reasoning vector during inference. This establishes a trade-off between efficiency and performance. Furthermore, we identify key properties of reasoning vectors: robustness, characterized by high similarity before and after training, and generalizability, demonstrating applicability across base models, distilled models, RL-trained models, parameter-merged models, and mixed-thought models. These properties collectively guarantee the reliability and broad applicability of our approach.</abstract>
<identifier type="citekey">han-etal-2026-errv</identifier>
<location>
<url>https://aclanthology.org/2026.findings-acl.1425/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>28557</start>
<end>28570</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T ERRV: Eliciting Efficient Reasoning through Reasoning Vectors for Policy Optimization in Large Language Models
%A Han, Zhuowen
%A Yang, Lei
%A Jin, Renren
%A Shi, Dan
%A Sun, Chenxi
%A Xiong, Deyi
%Y Liakata, Maria
%Y Moreira, Viviane P.
%Y Zhang, Jiajun
%Y Jurgens, David
%S Findings of the Association for Computational Linguistics: ACL 2026
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, United States
%@ 979-8-89176-395-1
%F han-etal-2026-errv
%X Recently, large reasoning models have achieved impressive performance, but their lengthy reasoning processes incur substantial inference overhead. To mitigate this issue, we propose the concept of reasoning vectors, representations extracted from the model’s hidden states, which can guide the model towards generating more concise and accurate responses. Building upon this, we present ERRV, a training framework that elicits efficient reasoning through reasoning vectors, which enables the model to generate high-quality responses during reinforcement learning. By performing targeted policy optimization on both accuracy and length objectives, ERRV effectively activates the model’s latent capability for efficient reasoning. Our experiments demonstrate that after training with ERRV, the model achieves approximately 30% reduction in reasoning length while maintaining stable accuracy, without guidance from the reasoning vector during inference. This establishes a trade-off between efficiency and performance. Furthermore, we identify key properties of reasoning vectors: robustness, characterized by high similarity before and after training, and generalizability, demonstrating applicability across base models, distilled models, RL-trained models, parameter-merged models, and mixed-thought models. These properties collectively guarantee the reliability and broad applicability of our approach.
%U https://aclanthology.org/2026.findings-acl.1425/
%P 28557-28570
Markdown (Informal)
[ERRV: Eliciting Efficient Reasoning through Reasoning Vectors for Policy Optimization in Large Language Models](https://aclanthology.org/2026.findings-acl.1425/) (Han et al., Findings 2026)
ACL