AnyMAC: Cascading Flexible Multi-Agent Collaboration via Next-Agent Prediction

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Wang, Song, Tan, Zhen, Chen, Zihan, Zhou, Shuang, Chen, Tianlong, Li, Jundong
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866917054092673024
author Wang, Song
Tan, Zhen
Chen, Zihan
Zhou, Shuang
Chen, Tianlong
Li, Jundong
author_facet Wang, Song
Tan, Zhen
Chen, Zihan
Zhou, Shuang
Chen, Tianlong
Li, Jundong
contents Recent progress in large language model (LLM)-based multi-agent collaboration highlights the power of structured communication in enabling collective intelligence. However, existing methods largely rely on static or graph-based inter-agent topologies, lacking the potential adaptability and flexibility in communication. In this work, we propose a new framework that rethinks multi-agent coordination through a sequential structure rather than a graph structure, offering a significantly larger topology space for multi-agent communication. Our method focuses on two key directions: (1) Next-Agent Prediction, which selects the most suitable agent role at each step, and (2) Next-Context Selection (NCS), which enables each agent to selectively access relevant information from any previous step. Together, these components construct task-adaptive communication pipelines that support both role flexibility and global information flow. Extensive evaluations across multiple benchmarks demonstrate that our approach achieves superior performance while substantially reducing communication overhead.
format Preprint
id arxiv_https___arxiv_org_abs_2506_17784
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle AnyMAC: Cascading Flexible Multi-Agent Collaboration via Next-Agent Prediction
Wang, Song
Tan, Zhen
Chen, Zihan
Zhou, Shuang
Chen, Tianlong
Li, Jundong
Artificial Intelligence
Recent progress in large language model (LLM)-based multi-agent collaboration highlights the power of structured communication in enabling collective intelligence. However, existing methods largely rely on static or graph-based inter-agent topologies, lacking the potential adaptability and flexibility in communication. In this work, we propose a new framework that rethinks multi-agent coordination through a sequential structure rather than a graph structure, offering a significantly larger topology space for multi-agent communication. Our method focuses on two key directions: (1) Next-Agent Prediction, which selects the most suitable agent role at each step, and (2) Next-Context Selection (NCS), which enables each agent to selectively access relevant information from any previous step. Together, these components construct task-adaptive communication pipelines that support both role flexibility and global information flow. Extensive evaluations across multiple benchmarks demonstrate that our approach achieves superior performance while substantially reducing communication overhead.
title AnyMAC: Cascading Flexible Multi-Agent Collaboration via Next-Agent Prediction
topic Artificial Intelligence
url https://arxiv.org/abs/2506.17784