@inproceedings{wu-etal-2026-st,
title = "{ST}-{EVO}: Towards Generative Spatio-Temporal Evolution of Multi-Agent Communication Topologies",
author = "Wu, Xingjian and
Liu, Xvyuan and
Lu, Junkai and
Wang, Siyuan and
Qiu, Xiangfei and
Shu, Yang and
Hu, Jilin and
Guo, Chenjuan and
Yang, Bin",
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.223/",
pages = "5156--5181",
ISBN = "979-8-89176-459-0",
abstract = "LLM-powered Multi-Agent Systems (MAS) have emerged as an effective approach towards collaborative intelligence, and have attracted wide research interests. Among them, ``self-evolving'' MAS, treated as a more flexible and powerful technical route, can construct task-adaptive workflows or communication topologies, instead of relying on a predefined static structure template. Current self-evolving MAS mainly focus on Spatial Evolving or Temporal Evolving paradigm, which only considers the single dimension of evolution and does not fully incentivize LLMs' collaborative capability. In this work, we start from a novel Spatio-Temporal perspective by proposing ST-EVO which supports dialogue-wise communication scheduling with a compact yet powerful flow-matching based Scheduler. To make precise Spatio-Temporal scheduling, the Scheduler can also perceive the uncertainty of MAS, and possesses self-feedback ability to learn from accumulated experience. Extensive experiments on nine benchmarks demonstrate the state-of-the-art performance of ST-EVO, achieving about 5{\%}{--}25{\%} accuracy improvement."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="wu-etal-2026-st">
<titleInfo>
<title>ST-EVO: Towards Generative Spatio-Temporal Evolution of Multi-Agent Communication Topologies</title>
</titleInfo>
<name type="personal">
<namePart type="given">Xingjian</namePart>
<namePart type="family">Wu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xvyuan</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Junkai</namePart>
<namePart type="family">Lu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Siyuan</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiangfei</namePart>
<namePart type="family">Qiu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yang</namePart>
<namePart type="family">Shu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jilin</namePart>
<namePart type="family">Hu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chenjuan</namePart>
<namePart type="family">Guo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Bin</namePart>
<namePart type="family">Yang</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>LLM-powered Multi-Agent Systems (MAS) have emerged as an effective approach towards collaborative intelligence, and have attracted wide research interests. Among them, “self-evolving” MAS, treated as a more flexible and powerful technical route, can construct task-adaptive workflows or communication topologies, instead of relying on a predefined static structure template. Current self-evolving MAS mainly focus on Spatial Evolving or Temporal Evolving paradigm, which only considers the single dimension of evolution and does not fully incentivize LLMs’ collaborative capability. In this work, we start from a novel Spatio-Temporal perspective by proposing ST-EVO which supports dialogue-wise communication scheduling with a compact yet powerful flow-matching based Scheduler. To make precise Spatio-Temporal scheduling, the Scheduler can also perceive the uncertainty of MAS, and possesses self-feedback ability to learn from accumulated experience. Extensive experiments on nine benchmarks demonstrate the state-of-the-art performance of ST-EVO, achieving about 5%–25% accuracy improvement.</abstract>
<identifier type="citekey">wu-etal-2026-st</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.223/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>5156</start>
<end>5181</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T ST-EVO: Towards Generative Spatio-Temporal Evolution of Multi-Agent Communication Topologies
%A Wu, Xingjian
%A Liu, Xvyuan
%A Lu, Junkai
%A Wang, Siyuan
%A Qiu, Xiangfei
%A Shu, Yang
%A Hu, Jilin
%A Guo, Chenjuan
%A Yang, Bin
%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 wu-etal-2026-st
%X LLM-powered Multi-Agent Systems (MAS) have emerged as an effective approach towards collaborative intelligence, and have attracted wide research interests. Among them, “self-evolving” MAS, treated as a more flexible and powerful technical route, can construct task-adaptive workflows or communication topologies, instead of relying on a predefined static structure template. Current self-evolving MAS mainly focus on Spatial Evolving or Temporal Evolving paradigm, which only considers the single dimension of evolution and does not fully incentivize LLMs’ collaborative capability. In this work, we start from a novel Spatio-Temporal perspective by proposing ST-EVO which supports dialogue-wise communication scheduling with a compact yet powerful flow-matching based Scheduler. To make precise Spatio-Temporal scheduling, the Scheduler can also perceive the uncertainty of MAS, and possesses self-feedback ability to learn from accumulated experience. Extensive experiments on nine benchmarks demonstrate the state-of-the-art performance of ST-EVO, achieving about 5%–25% accuracy improvement.
%U https://aclanthology.org/2026.emnlp-main.223/
%P 5156-5181
Markdown (Informal)
[ST-EVO: Towards Generative Spatio-Temporal Evolution of Multi-Agent Communication Topologies](https://aclanthology.org/2026.emnlp-main.223/) (Wu et al., EMNLP 2026)
ACL
- Xingjian Wu, Xvyuan Liu, Junkai Lu, Siyuan Wang, Xiangfei Qiu, Yang Shu, Jilin Hu, Chenjuan Guo, and Bin Yang. 2026. ST-EVO: Towards Generative Spatio-Temporal Evolution of Multi-Agent Communication Topologies. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 5156–5181, Budapest, Hungary. Association for Computational Linguistics.