CARD: Towards Conditional Design of Multi-agent Topological Structures

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Wu, Tongtong, Li, Yanming, Tang, Ziye, Jiang, Chen, Luo, Linhao, Qi, Guilin, Pan, Shirui, Haffari, Gholamreza
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866918363990589440
author Wu, Tongtong
Li, Yanming
Tang, Ziye
Jiang, Chen
Luo, Linhao
Qi, Guilin
Pan, Shirui
Haffari, Gholamreza
author_facet Wu, Tongtong
Li, Yanming
Tang, Ziye
Jiang, Chen
Luo, Linhao
Qi, Guilin
Pan, Shirui
Haffari, Gholamreza
contents Large language model (LLM)-based multi-agent systems have shown strong capabilities in tasks such as code generation and collaborative reasoning. However, the effectiveness and robustness of these systems critically depend on their communication topology, which is often fixed or statically learned, ignoring real-world dynamics such as model upgrades, API (or tool) changes, or knowledge source variability. To address this limitation, we propose CARD (Conditional Agentic Graph Designer), a conditional graph-generation framework that instantiates AMACP, a protocol for adaptive multi-agent communication. CARD explicitly incorporates dynamic environmental signals into graph construction, enabling topology adaptation at both training and runtime. Through a conditional variational graph encoder and environment-aware optimization, CARD produces communication structures that are both effective and resilient to shifts in model capability or resource availability. Empirical results on HumanEval, MATH, and MMLU demonstrate that CARD consistently outperforms static and prompt-based baselines, achieving higher accuracy and robustness across diverse conditions. The source code is available at: https://github.com/Warma10032/CARD.
format Preprint
id arxiv_https___arxiv_org_abs_2603_01089
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle CARD: Towards Conditional Design of Multi-agent Topological Structures
Wu, Tongtong
Li, Yanming
Tang, Ziye
Jiang, Chen
Luo, Linhao
Qi, Guilin
Pan, Shirui
Haffari, Gholamreza
Computation and Language
Machine Learning
Large language model (LLM)-based multi-agent systems have shown strong capabilities in tasks such as code generation and collaborative reasoning. However, the effectiveness and robustness of these systems critically depend on their communication topology, which is often fixed or statically learned, ignoring real-world dynamics such as model upgrades, API (or tool) changes, or knowledge source variability. To address this limitation, we propose CARD (Conditional Agentic Graph Designer), a conditional graph-generation framework that instantiates AMACP, a protocol for adaptive multi-agent communication. CARD explicitly incorporates dynamic environmental signals into graph construction, enabling topology adaptation at both training and runtime. Through a conditional variational graph encoder and environment-aware optimization, CARD produces communication structures that are both effective and resilient to shifts in model capability or resource availability. Empirical results on HumanEval, MATH, and MMLU demonstrate that CARD consistently outperforms static and prompt-based baselines, achieving higher accuracy and robustness across diverse conditions. The source code is available at: https://github.com/Warma10032/CARD.
title CARD: Towards Conditional Design of Multi-agent Topological Structures
topic Computation and Language
Machine Learning
url https://arxiv.org/abs/2603.01089