arXiv cs.CL· Yufeng Li, Zaijing Li, Fuchen Zhang, Zongzong Wu, Ming Zhao, Fengxiao Tang·· 13 小时前AI 评分31
ThreadRAG：用图拓扑与能量驱动寻路提升 RAG 上下文一致性
Graph Topology Meets Energy-Driven Pathfinding: Preserving Contextual Consistency in Retrieval-Augmented Generation
AI 导读
ThreadRAG 是一个基于语义拓扑的 RAG 框架，通过 Semantic Topology Mapper 将检索文档组织为上下文完整的 Semantic Nodes 并构建 Semantic Topology Graph，再用 Energy-Driven Pathfinding 经能量扩散定位查询相关子图并线性化为文本证据路径。
正文 · 原文
Abstract:Retrieval-augmented generation (RAG) frequently integrates knowledge graphs (KGs) for complex multi-hop question answering. However, their performance is increasingly bottlenecked by the knowledge structuring and reasoning processes. (i) Severe semantic loss: Distilling continuous texts into discrete symbols strips away contextual nuances and narrative coherence. (ii) Pronounced modality mismatch: Fragmented graph structures clash with the continuous texts LLMs are pre-trained on, while rewriting them into natural language introduces generative noise and omits information. To address these limitations, we propose ThreadRAG, a semantic-topology-based RAG framework for context-preserving knowledge structuring and modality-native evidence reasoning. ThreadRAG introduces a Semantic Topology Mapper (STM), which organizes retrieved documents into context-intact Semantic Nodes and constructs a Semantic Topology Graph (STG) through narrative, semantic, and lexical edges. It further incorporates Energy-Driven Pathfinding (EDP), which identifies query-relevant subgraphs via energy diffusion and deterministically linearizes them into coherent textual evidence paths. Experiments on three multi-hop QA benchmarks demonstrate that ThreadRAG consistently outperforms strong baselines. Notably, it achieves a 14.4% EM improvement over text-centric methods on HotpotQA and surpasses the best baseline by 5.8% EM on 2WikiMultihop, while reducing token cost to 23.3% of the strongest baseline.
| Comments: | Accepted to Findings of the Association for Computational Linguistics: EMNLP 2026 |
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.29933 [cs.CL] |
| (or arXiv:2610.29933v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.29933 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Yufeng Li [view email]
[v1]
Wed, 21 Oct 2026 16:05:44 UTC (549 KB)
来源：arXiv cs.CL · arxiv.org