Enregistré dans:
Détails bibliographiques
Auteurs principaux: Wei, Chunyu, Hu, Wenji, Hao, Xingjia, Wang, Xin, Yang, Yifan, Chen, Yueguo, Tian, Yang, Wang, Yunhai
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:https://arxiv.org/abs/2511.00457
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866908637535928320
author Wei, Chunyu
Hu, Wenji
Hao, Xingjia
Wang, Xin
Yang, Yifan
Chen, Yueguo
Tian, Yang
Wang, Yunhai
author_facet Wei, Chunyu
Hu, Wenji
Hao, Xingjia
Wang, Xin
Yang, Yifan
Chen, Yueguo
Tian, Yang
Wang, Yunhai
contents Large Language Models (LLMs) face significant limitations when applied to large-scale graphs, struggling with context constraints and inflexible reasoning. We present GraphChain, a framework that enables LLMs to analyze complex graphs through dynamic sequences of specialized tools, mimicking human exploratory intelligence. Our approach introduces two key innovations: (1) Progressive Graph Distillation, a reinforcement learning mechanism that generates optimized tool sequences balancing task relevance with information compression, and (2) Structure-aware Test-Time Adaptation, which efficiently tailors tool selection strategies to diverse graph topologies using spectral properties and lightweight adapters without costly retraining. Experiments show GraphChain significantly outperforms prior methods, enabling scalable and adaptive LLM-driven graph analysis.
format Preprint
id arxiv_https___arxiv_org_abs_2511_00457
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle GraphChain: Large Language Models for Large-scale Graph Analysis via Tool Chaining
Wei, Chunyu
Hu, Wenji
Hao, Xingjia
Wang, Xin
Yang, Yifan
Chen, Yueguo
Tian, Yang
Wang, Yunhai
Artificial Intelligence
Large Language Models (LLMs) face significant limitations when applied to large-scale graphs, struggling with context constraints and inflexible reasoning. We present GraphChain, a framework that enables LLMs to analyze complex graphs through dynamic sequences of specialized tools, mimicking human exploratory intelligence. Our approach introduces two key innovations: (1) Progressive Graph Distillation, a reinforcement learning mechanism that generates optimized tool sequences balancing task relevance with information compression, and (2) Structure-aware Test-Time Adaptation, which efficiently tailors tool selection strategies to diverse graph topologies using spectral properties and lightweight adapters without costly retraining. Experiments show GraphChain significantly outperforms prior methods, enabling scalable and adaptive LLM-driven graph analysis.
title GraphChain: Large Language Models for Large-scale Graph Analysis via Tool Chaining
topic Artificial Intelligence
url https://arxiv.org/abs/2511.00457