GraphAgent
LLM-based Agents - Generator, Reasonor - LLM4Graph
About The Project
The GraphAgent Framework is a platform built on LLM-based agents, designed to leverage the capabilities of LLMs for various graph data tasks. With its parallel acceleration capabilities, GraphAgent effectively meets the challenges posed by large-scale graph data. Currently, the framework comprises two modules: graph generation and graph reasoning. In the future, we plan to further enhance the GraphAgent framework.
This page is a collection of our work on GraphAgent.
Exploring the Potential of Large Language Models as Predictors in Dynamic Text-Attributed Graphs (Lei et al., 2025)
Citation
@misc{lei2025exploring,
title={Exploring the Potential of Large Language Models as Predictors in Dynamic Text-Attributed Graphs},
author={Runlin Lei and Jiarui Ji and Haipeng Ding and Lu Yi and Zhewei Wei and Yongchao Liu and Chuntao Hong},
year={2025},
eprint={2503.03258},
archivePrefix={arXiv},
primaryClass={cs.LG},
url={https://arxiv.org/abs/2503.03258},
}
Scalable and Accurate Graph Reasoning with LLM-based Multi-Agents (Hu et al., 2024)
Citation
@article{hu2024scalable,
title={Scalable and Accurate Graph Reasoning with LLM-based Multi-Agents},
author={Hu, Yuwei and Lei, Runlin and Huang, Xinyi and Wei, Zhewei and Liu, Yongchao},
journal={arXiv preprint arXiv:2410.05130},
year={2024}
}
Dynamic and Textual Graph Generation Via Large-Scale LLM-based Agent Simulation (Ji et al., 2025)
Citation
@preprint{ji2025llm,
title={LLM-Based Multi-Agent Systems are Scalable Graph Generative Models},
author={Jiarui Ji and Runlin Lei and Jialing Bi and Zhewei Wei and Xu Chen and Yankai Lin and Xuchen Pan and Yaliang Li and Bolin Ding},
year={2025},
eprint={2410.09824},
archivePrefix={arXiv},
primaryClass={cs.CL},
url={https://arxiv.org/abs/2410.09824},
}