@inproceedings{zhao-etal-2026-bias-aware,
title = "Bias-Aware Orchestration: Role-Specialized Multi-Agent Collaboration of Heterogeneous {LLM}s",
author = "Zhao, Yuanpei and
Zhang, Haowei and
Qiao, Ziyue and
Lv, Jiancheng and
li, Mao and
Ju, Wei",
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.2055/",
pages = "43768--43782",
ISBN = "979-8-89176-462-0",
abstract = "Coordinating heterogeneous large language models (LLMs) can improve instruction following, yet current multi-agent systems often depend on manual role choices or repeated calls to one backbone. This makes it difficult to separate complementary model behavior from model bias and verbosity. We introduce Role-Specialized Multi-Agent Collaboration (RS-MA), a prompting-only framework that assigns candidate LLMs to four interpretable roles: Judge, Summarizer, Ideator, and Reasoner. RS-MA estimates role fit with audition prompts, aggregates role-specific rankings with Borda scoring, and uses the selected assignment as an inspectable role map. A refinement phase records judgment gaps and targeted edits before synthesis. On AlpacaEval 2.0, audition-selected RS-MA reaches an 88.86{\%} length-controlled win rate, and a validated role map reaches 90.72{\%}. A full 805-example ablation shows that single-model reuse attains high raw preference through much longer outputs, whereas the validated map improves length-controlled quality with moderate length growth. MT-Bench and FLASK further indicate preserved dialogue quality and balanced fine-grained skills. The results support explicit role assignment as a practical mechanism for combining heterogeneous LLMs when quality, completeness, and inspectable evidence are jointly required."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zhao-etal-2026-bias-aware">
<titleInfo>
<title>Bias-Aware Orchestration: Role-Specialized Multi-Agent Collaboration of Heterogeneous LLMs</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yuanpei</namePart>
<namePart type="family">Zhao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Haowei</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ziyue</namePart>
<namePart type="family">Qiao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiancheng</namePart>
<namePart type="family">Lv</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mao</namePart>
<namePart type="family">li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wei</namePart>
<namePart type="family">Ju</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>Coordinating heterogeneous large language models (LLMs) can improve instruction following, yet current multi-agent systems often depend on manual role choices or repeated calls to one backbone. This makes it difficult to separate complementary model behavior from model bias and verbosity. We introduce Role-Specialized Multi-Agent Collaboration (RS-MA), a prompting-only framework that assigns candidate LLMs to four interpretable roles: Judge, Summarizer, Ideator, and Reasoner. RS-MA estimates role fit with audition prompts, aggregates role-specific rankings with Borda scoring, and uses the selected assignment as an inspectable role map. A refinement phase records judgment gaps and targeted edits before synthesis. On AlpacaEval 2.0, audition-selected RS-MA reaches an 88.86% length-controlled win rate, and a validated role map reaches 90.72%. A full 805-example ablation shows that single-model reuse attains high raw preference through much longer outputs, whereas the validated map improves length-controlled quality with moderate length growth. MT-Bench and FLASK further indicate preserved dialogue quality and balanced fine-grained skills. The results support explicit role assignment as a practical mechanism for combining heterogeneous LLMs when quality, completeness, and inspectable evidence are jointly required.</abstract>
<identifier type="citekey">zhao-etal-2026-bias-aware</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.2055/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>43768</start>
<end>43782</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Bias-Aware Orchestration: Role-Specialized Multi-Agent Collaboration of Heterogeneous LLMs
%A Zhao, Yuanpei
%A Zhang, Haowei
%A Qiao, Ziyue
%A Lv, Jiancheng
%A li, Mao
%A Ju, Wei
%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 zhao-etal-2026-bias-aware
%X Coordinating heterogeneous large language models (LLMs) can improve instruction following, yet current multi-agent systems often depend on manual role choices or repeated calls to one backbone. This makes it difficult to separate complementary model behavior from model bias and verbosity. We introduce Role-Specialized Multi-Agent Collaboration (RS-MA), a prompting-only framework that assigns candidate LLMs to four interpretable roles: Judge, Summarizer, Ideator, and Reasoner. RS-MA estimates role fit with audition prompts, aggregates role-specific rankings with Borda scoring, and uses the selected assignment as an inspectable role map. A refinement phase records judgment gaps and targeted edits before synthesis. On AlpacaEval 2.0, audition-selected RS-MA reaches an 88.86% length-controlled win rate, and a validated role map reaches 90.72%. A full 805-example ablation shows that single-model reuse attains high raw preference through much longer outputs, whereas the validated map improves length-controlled quality with moderate length growth. MT-Bench and FLASK further indicate preserved dialogue quality and balanced fine-grained skills. The results support explicit role assignment as a practical mechanism for combining heterogeneous LLMs when quality, completeness, and inspectable evidence are jointly required.
%U https://aclanthology.org/2026.findings-emnlp.2055/
%P 43768-43782
Markdown (Informal)
[Bias-Aware Orchestration: Role-Specialized Multi-Agent Collaboration of Heterogeneous LLMs](https://aclanthology.org/2026.findings-emnlp.2055/) (Zhao et al., Findings 2026)
ACL