Guardado en:
Detalles Bibliográficos
Autores principales: Tang, Jinzhou, Zhang, Jusheng, Lv, Qinhan, Liu, Sidi, Yang, Jing, Tang, Chengpei, Wang, Keze
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:https://arxiv.org/abs/2509.00189
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866912559859236864
author Tang, Jinzhou
Zhang, Jusheng
Lv, Qinhan
Liu, Sidi
Yang, Jing
Tang, Chengpei
Wang, Keze
author_facet Tang, Jinzhou
Zhang, Jusheng
Lv, Qinhan
Liu, Sidi
Yang, Jing
Tang, Chengpei
Wang, Keze
contents Autonomous agents play a crucial role in advancing Artificial General Intelligence, enabling problem decomposition and tool orchestration through Large Language Models (LLMs). However, existing paradigms face a critical trade-off. On one hand, reusable fixed workflows require manual reconfiguration upon environmental changes; on the other hand, flexible reactive loops fail to distill reasoning progress into transferable structures. We introduce Hierarchical Variable Agent (HiVA), a novel framework modeling agentic workflows as self-organized graphs with the Semantic-Topological Evolution (STEV) algorithm, which optimizes hybrid semantic-topological spaces using textual gradients as discrete-domain surrogates for backpropagation. The iterative process comprises Multi-Armed Bandit-infused forward routing, diagnostic gradient generation from environmental feedback, and coordinated updates that co-evolve individual semantics and topology for collective optimization in unknown environments. Experiments on dialogue, coding, Long-context Q&A, mathematical, and agentic benchmarks demonstrate improvements of 5-10% in task accuracy and enhanced resource efficiency over existing baselines, establishing HiVA's effectiveness in autonomous task execution.
format Preprint
id arxiv_https___arxiv_org_abs_2509_00189
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle HiVA: Self-organized Hierarchical Variable Agent via Goal-driven Semantic-Topological Evolution
Tang, Jinzhou
Zhang, Jusheng
Lv, Qinhan
Liu, Sidi
Yang, Jing
Tang, Chengpei
Wang, Keze
Artificial Intelligence
Multiagent Systems
Autonomous agents play a crucial role in advancing Artificial General Intelligence, enabling problem decomposition and tool orchestration through Large Language Models (LLMs). However, existing paradigms face a critical trade-off. On one hand, reusable fixed workflows require manual reconfiguration upon environmental changes; on the other hand, flexible reactive loops fail to distill reasoning progress into transferable structures. We introduce Hierarchical Variable Agent (HiVA), a novel framework modeling agentic workflows as self-organized graphs with the Semantic-Topological Evolution (STEV) algorithm, which optimizes hybrid semantic-topological spaces using textual gradients as discrete-domain surrogates for backpropagation. The iterative process comprises Multi-Armed Bandit-infused forward routing, diagnostic gradient generation from environmental feedback, and coordinated updates that co-evolve individual semantics and topology for collective optimization in unknown environments. Experiments on dialogue, coding, Long-context Q&A, mathematical, and agentic benchmarks demonstrate improvements of 5-10% in task accuracy and enhanced resource efficiency over existing baselines, establishing HiVA's effectiveness in autonomous task execution.
title HiVA: Self-organized Hierarchical Variable Agent via Goal-driven Semantic-Topological Evolution
topic Artificial Intelligence
Multiagent Systems
url https://arxiv.org/abs/2509.00189