GraphScout: Empowering Large Language Models with Intrinsic Exploration Ability for Agentic Graph Reasoning
Open Access
- 8 August 2026
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 6258-6269
Abstract
Knowledge graphs provide structured and reliable information for many real-world applications, motivating increasing interest in combining large language models (LLMs) with graph-based retrieval to improve factual grounding. Recent Graph-based Retrieval-Augmented Generation (GraphRAG) methods therefore introduce iterative interaction between LLMs and knowledge graphs to enhance reasoning capability. However, existing approaches typically depend on manually designed guidance and interact with knowledge graphs through a limited set of predefined tools, which substantially constrains graph exploration. To address these limitations, we propose GraphScout, a training-centric agentic graph reasoning framework equipped with more flexible graph exploration tools. GraphScout enables models to autonomously interact with knowledge graphs to synthesize structured training data which are then used to post-train LLMs, thereby internalizing agentic graph reasoning ability without laborious manual annotation or task curation. Extensive experiments across five knowledge-graph domains show that a small model (e.g., Qwen3-4B) augmented with GraphScout outperforms baseline methods built on leading LLMs (e.g., Qwen-Max) by an average of 16.7% while requiring significantly fewer inference tokens. Moreover, GraphScout exhibits robust cross-domain transfer performance. Our code is publicly available https://github.com/Ying-Yuchen/_GraphScout_.Keywords
All Related Versions
This publication has 14 references indexed in Scilit:
- Neural Graph Navigation for Intelligent Subgraph MatchingProceedings of the AAAI Conference on Artificial Intelligence, 2026
- Association Pattern-enhanced Molecular Representation LearningProceedings of the AAAI Conference on Artificial Intelligence, 2025
- Language Models-enhanced Semantic Topology Representation Learning For Temporal Knowledge Graph Extrapolation2024
- Spatiotemporal-Augmented Graph Neural Networks for Human Mobility SimulationIEEE Transactions on Knowledge and Data Engineering, 2024
- Exploring the Potential of Large Language Models (LLMs)in Learning on GraphsACM SIGKDD Explorations Newsletter, 2024
- Benchmarking Large Language Models in Retrieval-Augmented GenerationProceedings of the AAAI Conference on Artificial Intelligence, 2024
- Temporal Aggregation and Propagation Graph Neural Networks for Dynamic RepresentationIEEE Transactions on Knowledge and Data Engineering, 2023
- Retrieval Augmentation Reduces Hallucination in Conversation2021
- Characteristic Functions on Graphs: Birds of a Feather, from Statistical Descriptors to Parametric Models2020
- Real-time intelligent big data processing: technology, platform, and applicationsScience China Information Sciences, 2019