@inproceedings{zhou-etal-2026-orchdebate,
title = "{O}rch{D}ebate: Adaptive Orchestration for Multi-Agent Debate via Self-Evolving Skills",
author = "Zhou, Kaiyu and
Zheng, Yongsen and
Zhang, Xuanye and
Lam, Kwok-Yan",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-main.2078/",
pages = "47331--47350",
ISBN = "979-8-89176-459-0",
abstract = "Multi-Agent Debate (MAD) has emerged as a promising paradigm for eliciting stronger inference-time reasoning, yet most existing systems still use fixed interaction protocols across instances. Such static designs can suffer from anchoring bias, overthinking on easy problems, and rigid coordination across tasks of varying difficulty. We present OrchDebate, a framework that recasts multi-agent debate as structured-state orchestration over reusable debate-control skills. Instead of following a pre-defined template, OrchDebate uses an orchestrator to examine the current debate state and select the next coordination action, such as introducing agents under different visibility settings or terminating deliberation with an adaptive readout rule. This skill-based view makes debate control modular and state-dependent. The orchestrator operates through a structured control interface rather than directly generating reasoning content, helping reduce reliance on early answers. OrchDebate further maintains a self-evolving skill bank whose reusable realizations are refined through outcome-guided updates on an evolution stream, without additional external training data. Across diverse benchmarks, OrchDebate delivers a better accuracy-cost trade-off than representative fixed and adaptive debate methods by allocating debate compute more selectively."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zhou-etal-2026-orchdebate">
<titleInfo>
<title>OrchDebate: Adaptive Orchestration for Multi-Agent Debate via Self-Evolving Skills</title>
</titleInfo>
<name type="personal">
<namePart type="given">Kaiyu</namePart>
<namePart type="family">Zhou</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yongsen</namePart>
<namePart type="family">Zheng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xuanye</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kwok-Yan</namePart>
<namePart type="family">Lam</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>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</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">979-8-89176-459-0</identifier>
</relatedItem>
<abstract>Multi-Agent Debate (MAD) has emerged as a promising paradigm for eliciting stronger inference-time reasoning, yet most existing systems still use fixed interaction protocols across instances. Such static designs can suffer from anchoring bias, overthinking on easy problems, and rigid coordination across tasks of varying difficulty. We present OrchDebate, a framework that recasts multi-agent debate as structured-state orchestration over reusable debate-control skills. Instead of following a pre-defined template, OrchDebate uses an orchestrator to examine the current debate state and select the next coordination action, such as introducing agents under different visibility settings or terminating deliberation with an adaptive readout rule. This skill-based view makes debate control modular and state-dependent. The orchestrator operates through a structured control interface rather than directly generating reasoning content, helping reduce reliance on early answers. OrchDebate further maintains a self-evolving skill bank whose reusable realizations are refined through outcome-guided updates on an evolution stream, without additional external training data. Across diverse benchmarks, OrchDebate delivers a better accuracy-cost trade-off than representative fixed and adaptive debate methods by allocating debate compute more selectively.</abstract>
<identifier type="citekey">zhou-etal-2026-orchdebate</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.2078/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>47331</start>
<end>47350</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T OrchDebate: Adaptive Orchestration for Multi-Agent Debate via Self-Evolving Skills
%A Zhou, Kaiyu
%A Zheng, Yongsen
%A Zhang, Xuanye
%A Lam, Kwok-Yan
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F zhou-etal-2026-orchdebate
%X Multi-Agent Debate (MAD) has emerged as a promising paradigm for eliciting stronger inference-time reasoning, yet most existing systems still use fixed interaction protocols across instances. Such static designs can suffer from anchoring bias, overthinking on easy problems, and rigid coordination across tasks of varying difficulty. We present OrchDebate, a framework that recasts multi-agent debate as structured-state orchestration over reusable debate-control skills. Instead of following a pre-defined template, OrchDebate uses an orchestrator to examine the current debate state and select the next coordination action, such as introducing agents under different visibility settings or terminating deliberation with an adaptive readout rule. This skill-based view makes debate control modular and state-dependent. The orchestrator operates through a structured control interface rather than directly generating reasoning content, helping reduce reliance on early answers. OrchDebate further maintains a self-evolving skill bank whose reusable realizations are refined through outcome-guided updates on an evolution stream, without additional external training data. Across diverse benchmarks, OrchDebate delivers a better accuracy-cost trade-off than representative fixed and adaptive debate methods by allocating debate compute more selectively.
%U https://aclanthology.org/2026.emnlp-main.2078/
%P 47331-47350
Markdown (Informal)
[OrchDebate: Adaptive Orchestration for Multi-Agent Debate via Self-Evolving Skills](https://aclanthology.org/2026.emnlp-main.2078/) (Zhou et al., EMNLP 2026)
ACL