@inproceedings{wang-etal-2026-role,
title = "Role-Agent: Bootstrapping {LLM} Agents via Dual-Role Evolution",
author = "Wang, Xucong and
Ma, Ziyu and
Yang, Shidong and
Huang, Tongwen and
Wang, Pengkun and
Wang, Yong and
Chu, Xiangxiang",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.208/",
pages = "4360--4379",
ISBN = "979-8-89176-462-0",
abstract = "Although Large Language Model (LLM) agents have demonstrated remarkable performance on complex, real-world tasks, the significant bottleneck of execution inefficiencies and static environmental feedback hinders agents' effective generalization to open-ended scenarios. To address these limitations, this paper introduces Role-Agent, a novel framework that repurposes one LLM to concurrently function as the agent and the environment, thereby enabling a bootstrapped co-evolution. At the core of Role-Agent are two synergistic components: World-In-Agent (WIA) and Agent-In-World (AIW). In WIA where the LLM acts as the agent, we define the reward for an action as the alignment between the agent{'}s prediction of the next several states and the actual state. These rewards are subsequently grouped by state and optimized, equipping the agent with predictive reasoning capabilities regarding the environment. In AIW where the LLM provides environmental adjustments, we require it to analyze the typical failure modes from failed trajectories, then searches for tasks that exhibit similar failure modes, thereby altering the data distribution to facilitate learning. Extensive experiments conducted on the various benchmarks demonstrate the substantial performance improvements achieved by Role-Agent, yielding an average performance gain of over 4{\%}."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="wang-etal-2026-role">
<titleInfo>
<title>Role-Agent: Bootstrapping LLM Agents via Dual-Role Evolution</title>
</titleInfo>
<name type="personal">
<namePart type="given">Xucong</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ziyu</namePart>
<namePart type="family">Ma</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shidong</namePart>
<namePart type="family">Yang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tongwen</namePart>
<namePart type="family">Huang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Pengkun</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yong</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiangxiang</namePart>
<namePart type="family">Chu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>Although Large Language Model (LLM) agents have demonstrated remarkable performance on complex, real-world tasks, the significant bottleneck of execution inefficiencies and static environmental feedback hinders agents’ effective generalization to open-ended scenarios. To address these limitations, this paper introduces Role-Agent, a novel framework that repurposes one LLM to concurrently function as the agent and the environment, thereby enabling a bootstrapped co-evolution. At the core of Role-Agent are two synergistic components: World-In-Agent (WIA) and Agent-In-World (AIW). In WIA where the LLM acts as the agent, we define the reward for an action as the alignment between the agent’s prediction of the next several states and the actual state. These rewards are subsequently grouped by state and optimized, equipping the agent with predictive reasoning capabilities regarding the environment. In AIW where the LLM provides environmental adjustments, we require it to analyze the typical failure modes from failed trajectories, then searches for tasks that exhibit similar failure modes, thereby altering the data distribution to facilitate learning. Extensive experiments conducted on the various benchmarks demonstrate the substantial performance improvements achieved by Role-Agent, yielding an average performance gain of over 4%.</abstract>
<identifier type="citekey">wang-etal-2026-role</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.208/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>4360</start>
<end>4379</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Role-Agent: Bootstrapping LLM Agents via Dual-Role Evolution
%A Wang, Xucong
%A Ma, Ziyu
%A Yang, Shidong
%A Huang, Tongwen
%A Wang, Pengkun
%A Wang, Yong
%A Chu, Xiangxiang
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F wang-etal-2026-role
%X Although Large Language Model (LLM) agents have demonstrated remarkable performance on complex, real-world tasks, the significant bottleneck of execution inefficiencies and static environmental feedback hinders agents’ effective generalization to open-ended scenarios. To address these limitations, this paper introduces Role-Agent, a novel framework that repurposes one LLM to concurrently function as the agent and the environment, thereby enabling a bootstrapped co-evolution. At the core of Role-Agent are two synergistic components: World-In-Agent (WIA) and Agent-In-World (AIW). In WIA where the LLM acts as the agent, we define the reward for an action as the alignment between the agent’s prediction of the next several states and the actual state. These rewards are subsequently grouped by state and optimized, equipping the agent with predictive reasoning capabilities regarding the environment. In AIW where the LLM provides environmental adjustments, we require it to analyze the typical failure modes from failed trajectories, then searches for tasks that exhibit similar failure modes, thereby altering the data distribution to facilitate learning. Extensive experiments conducted on the various benchmarks demonstrate the substantial performance improvements achieved by Role-Agent, yielding an average performance gain of over 4%.
%U https://aclanthology.org/2026.findings-emnlp.208/
%P 4360-4379
Markdown (Informal)
[Role-Agent: Bootstrapping LLM Agents via Dual-Role Evolution](https://aclanthology.org/2026.findings-emnlp.208/) (Wang et al., Findings 2026)
ACL
- Xucong Wang, Ziyu Ma, Shidong Yang, Tongwen Huang, Pengkun Wang, Yong Wang, and Xiangxiang Chu. 2026. Role-Agent: Bootstrapping LLM Agents via Dual-Role Evolution. In Findings of the Association for Computational Linguistics: EMNLP 2026, pages 4360–4379, Budapest, Hungary. Association for Computational Linguistics.