Think-on-Graph 3.0: Efficient and Adaptive LLM Reasoning on Heterogeneous Graphs via Multi-Agent Dual-Evolving Context Retrieval

2025-09-25

AI Summary

Existing graph-structured RAG methods face bottlenecks in graph construction: manually curated knowledge graphs are costly to scale, while automated graph extraction heavily relies on LLM-based extractors—yielding suboptimal performance, especially with lightweight LLMs. Method: We propose ToG-3, the first framework featuring a heterogeneous graph index composed of chunks, triplets, and communities, coupled with a novel “query–subgraph” dual-evolution mechanism that enables dynamic, adaptive graph construction during inference. We further design MACER—a multi-agent collaborative architecture comprising Constructor, Retriever, Reflector, and Responder—that jointly integrates retrieval-augmented generation, dynamic graph construction, contextual evolution, and iterative reflection. Results: Experiments demonstrate that ToG-3 significantly outperforms baselines on both depth- and breadth-oriented reasoning benchmarks. Ablation studies confirm the effectiveness of each component, and ToG-3 maintains strong performance even when deployed with lightweight LLMs.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community
• Knowledge Representation and Reasoning: Computational Complexity of Reasoning

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Retrieval-Augmented Generation (RAG) and Graph-based RAG has become the important paradigm for enhancing Large Language Models (LLMs) with external knowledge. However, existing approaches face a fundamental trade-off. While graph-based methods are inherently dependent on high-quality graph structures, they face significant practical constraints: manually constructed knowledge graphs are prohibitively expensive to scale, while automatically extracted graphs from corpora are limited by the performance of the underlying LLM extractors, especially when using smaller, local-deployed models. This paper presents Think-on-Graph 3.0 (ToG-3), a novel framework that introduces Multi-Agent Context Evolution and Retrieval (MACER) mechanism to overcome these limitations. Our core innovation is the dynamic construction and refinement of a Chunk-Triplets-Community heterogeneous graph index, which pioneeringly incorporates a dual-evolution mechanism of Evolving Query and Evolving Sub-Graph for precise evidence retrieval. This approach addresses a critical limitation of prior Graph-based RAG methods, which typically construct a static graph index in a single pass without adapting to the actual query. A multi-agent system, comprising Constructor, Retriever, Reflector, and Responser agents, collaboratively engages in an iterative process of evidence retrieval, answer generation, sufficiency reflection, and, crucially, evolving query and subgraph. This dual-evolving multi-agent system allows ToG-3 to adaptively build a targeted graph index during reasoning, mitigating the inherent drawbacks of static, one-time graph construction and enabling deep, precise reasoning even with lightweight LLMs. Extensive experiments demonstrate that ToG-3 outperforms compared baselines on both deep and broad reasoning benchmarks, and ablation studies confirm the efficacy of the components of MACER framework.

Problem

• Dynamic graph construction adapts to queries for precise evidence retrieval
• Overcomes static graph limitations in knowledge-enhanced language models
• Enables deep reasoning with lightweight models through multi-agent collaboration

Innovation

• Dual-evolving query and subgraph mechanism for retrieval
• Multi-agent system for iterative evidence retrieval and refinement
• Dynamic heterogeneous graph index construction during reasoning