Research map / Large language models / Retrieval and knowledge
LinearRAG: Linear Graph Retrieval Augmented Generation on Large-scale Corpora
Luyao Zhuang, Shengyuan Chen, Yilin Xiao, Huachi Zhou, Yujing Zhang, Hao Chen, Qinggang Zhang, Xiao Huang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add relevant entity activation via semantic bridging (ablation) | LinearRAG · four datasets | average of GPT-Acc and Contain-Acc | Improved |
| add global importance aggregation via personalized PageRank (ablation) | LinearRAG · four datasets | average of GPT-Acc and Contain-Acc | Improved |
| replace all-mpnet-base-v2 with e5-large-v2 | LinearRAG · HotpotQA | Contain-Acc: 64.30 -> 66.80 | Improved |
| replace all-mpnet-base-v2 with e5-large-v2 | LinearRAG · MuSiQue | Contain-Acc: 33.90 -> 31.70 | Hurt |
| replace all-mpnet-base-v2 with all-MiniLM-L6-v2 | LinearRAG · Medical | GPT-Acc: 63.72 -> 62.32 | Hurt |
| replace all-mpnet-base-v2 with bge-large-en-v1.5 | LinearRAG · Medical | GPT-Acc: 63.72 -> 64.45 | Improved |
| use LinearRAG instead of HippoRAG2 graph construction | HippoRAG2 · 2Wiki | GPT-Acc: 55.00 -> 63.70 | Improved |
| add graph retrieval (GFM-RAG) over vanilla RAG | Vanilla RAG top-5 · creative generation (Medical) | relevance: 58.73 -> 22.87 | Hurt |
| add graph retrieval (GFM-RAG) over vanilla RAG | Vanilla RAG top-5 · creative generation (Medical) | evidence recall: 44.88 -> 83.51 | Improved |
| use LinearRAG instead of GFM-RAG | GFM-RAG · complex reasoning (Medical) | relevance: 33.06 -> 81.58 | Improved |
| use LinearRAG instead of RAPTOR indexing | RAPTOR · ATLAS-Wiki 10M tokens | indexing time (s): 46430.96 -> 3084.38 | Improved |
| increase retrieved passages from top-1 to top-5 | vanilla RAG · MuSiQue | GPT-Acc: 21.10 -> 29.60 | Improved |
Later papers that compared with this method
2 later accepted papers report LinearRAG as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- GAM-RAG: Gain-Adaptive Memory for Evolving Retrieval in Retrieval-Augmented Generation (ICML 2026 · better on 4 of 4 shared results)
- Geometric Gain Graph: Zero-Token Graph Construction for Multi-Hop RAG (NeurIPS 2026 · better on 1 of 1 shared result)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- HippoRAG2 — HippoRAG: Neurobiologically Inspired Long-Term Memory for Large Language Models (NeurIPS 2024) · better on 9 of 9 shared results
- RAPTOR — RAPTOR: Recursive Abstractive Processing for Tree-Organized Retrieval (ICLR 2024) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Simple is Effective: The Roles of Graphs and Large Language Models in Knowledge-Graph-Based Retrieval-Augmented Generation (ICLR 2025 · similarity 0.92)
- A Systematic Exploration of Knowledge Graph Alignment with Large Language Models in Retrieval Augmented Generation (AAAI 2025 · similarity 0.92)
- Think-on-Graph 2.0: Deep and Faithful Large Language Model Reasoning with Knowledge-guided Retrieval Augmented Generation (ICLR 2025 · similarity 0.91)
- HyperGraphRAG: Retrieval-Augmented Generation via Hypergraph-Structured Knowledge Representation (NeurIPS 2025 · similarity 0.91)
- GFM-RAG: Graph Foundation Model for Retrieval Augmented Generation (NeurIPS 2025 · similarity 0.90)
- Can Knowledge-Graph-based Retrieval Augmented Generation Really Retrieve What You Need? (NeurIPS 2025 · similarity 0.90)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
