Saved in:
| Main Authors: | , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2509.24276 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866918476647497728 |
|---|---|
| author | Luo, Linhao Zhao, Zicheng Liu, Junnan Qiu, Zhangchi Dong, Junnan Panev, Serge Gong, Chen Vu, Thuy-Trang Haffari, Gholamreza Phung, Dinh Liew, Alan Wee-Chung Pan, Shirui |
| author_facet | Luo, Linhao Zhao, Zicheng Liu, Junnan Qiu, Zhangchi Dong, Junnan Panev, Serge Gong, Chen Vu, Thuy-Trang Haffari, Gholamreza Phung, Dinh Liew, Alan Wee-Chung Pan, Shirui |
| contents | Large language models (LLMs) excel at complex reasoning but remain limited by static and incomplete parametric knowledge. Retrieval-augmented generation (RAG) mitigates this by incorporating external knowledge, yet existing RAGs struggle with knowledge-intensive tasks due to fragmented information and weak modeling of knowledge structure. Graphs offer a natural way to model relationships within knowledge, but LLMs are inherently unstructured and cannot effectively reason over graph-structured data. Recent graph-enhanced RAG (GraphRAG) attempts to bridge this gap by constructing tailored graphs and enabling LLMs to reason on them. However, these methods often depend on ad-hoc graph designs, heuristic search, or costly agent pipelines, which hinder scalability and generalization. To address these challenges, we present G-reasoner, a unified framework that integrates graph and language foundation models for scalable reasoning over diverse graph-structured knowledge. Central to our approach is QuadGraph, a standardized four-layer abstraction that unifies heterogeneous knowledge sources into a common graph representation. Building on this, we introduce a 34M-parameter graph foundation model (GFM) that jointly captures graph topology and textual semantics, and is integrated with LLMs to enhance reasoning in downstream applications. To ensure scalability and efficiency, mixed-precision training and distributed message-passing are implemented to scale GFM with more GPUs. Extensive experiments on six benchmarks show that G-reasoner consistently outperforms state-of-the-art baselines, significantly enhances LLM reasoning, and achieves strong efficiency and cross-graph generalization. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_24276 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | G-reasoner: Foundation Models for Unified Reasoning over Graph-structured Knowledge Luo, Linhao Zhao, Zicheng Liu, Junnan Qiu, Zhangchi Dong, Junnan Panev, Serge Gong, Chen Vu, Thuy-Trang Haffari, Gholamreza Phung, Dinh Liew, Alan Wee-Chung Pan, Shirui Artificial Intelligence Large language models (LLMs) excel at complex reasoning but remain limited by static and incomplete parametric knowledge. Retrieval-augmented generation (RAG) mitigates this by incorporating external knowledge, yet existing RAGs struggle with knowledge-intensive tasks due to fragmented information and weak modeling of knowledge structure. Graphs offer a natural way to model relationships within knowledge, but LLMs are inherently unstructured and cannot effectively reason over graph-structured data. Recent graph-enhanced RAG (GraphRAG) attempts to bridge this gap by constructing tailored graphs and enabling LLMs to reason on them. However, these methods often depend on ad-hoc graph designs, heuristic search, or costly agent pipelines, which hinder scalability and generalization. To address these challenges, we present G-reasoner, a unified framework that integrates graph and language foundation models for scalable reasoning over diverse graph-structured knowledge. Central to our approach is QuadGraph, a standardized four-layer abstraction that unifies heterogeneous knowledge sources into a common graph representation. Building on this, we introduce a 34M-parameter graph foundation model (GFM) that jointly captures graph topology and textual semantics, and is integrated with LLMs to enhance reasoning in downstream applications. To ensure scalability and efficiency, mixed-precision training and distributed message-passing are implemented to scale GFM with more GPUs. Extensive experiments on six benchmarks show that G-reasoner consistently outperforms state-of-the-art baselines, significantly enhances LLM reasoning, and achieves strong efficiency and cross-graph generalization. |
| title | G-reasoner: Foundation Models for Unified Reasoning over Graph-structured Knowledge |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2509.24276 |