Research map / Large language models / Retrieval and knowledge
Token-Free Hierarchical Indexing for RAG beyond LLM-based Summarization
Yifan Wei, Dan Yuan, Xiaoyan Yu, Angsheng Li
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 self-query query enhancement (ablation) | SeRAG · 2WikiMultiHopQA | string accuracy: 63.6 -> 64.4 | Improved |
| add self-query query enhancement (ablation) | SeRAG · 2WikiMultiHopQA | LLM accuracy: 60.3 -> 61.5 | Improved |
| add logical and distance layers to the semantic graph (ablation) | SeRAG · 2WikiMultiHopQA | string accuracy: 61.3 -> 64.4 | Improved |
| add logical and distance layers to the semantic graph (ablation) | SeRAG · 2WikiMultiHopQA | LLM accuracy: 59.2 -> 61.5 | Improved |
| add structural-entropy hierarchical index over flat retrieval (ablation) | SeRAG · 2WikiMultiHopQA | string accuracy: 44.9 -> 64.4 | Improved |
| add structural-entropy hierarchical index over flat retrieval (ablation) | SeRAG · 2WikiMultiHopQA | LLM accuracy: 43.6 -> 61.5 | Improved |
| increase retrieved units k from 7 to 9 | SeRAG · 2WikiMultiHopQA | string accuracy: 67.1 -> 66.8 | Hurt |
| increase retrieved units k from 7 to 9 | SeRAG · 2WikiMultiHopQA | LLM accuracy: 63.9 -> 62.9 | Hurt |
| increase retrieved units k from 3 to 7 | SeRAG · MuSiQue | LLM accuracy: 32.1 -> 35.8 | Improved |
| increase retrieved units k from 3 to 7 | SeRAG · MuSiQue | string accuracy: 29.7 -> 31.8 | Improved |
| increase retrieved units k from 7 to 9 | SeRAG · MuSiQue | LLM accuracy: 35.8 -> 33.7 | Hurt |
| replace LLM summarization indexing (RAPTOR) with token-free vector consolidation | SeRAG vs RAPTOR · 2WikiMultiHopQA | prompt tokens (x10^6): 0.81 -> 0.38 | Improved |
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 11 of 12 shared results
- RAPTOR — RAPTOR: Recursive Abstractive Processing for Tree-Organized Retrieval (ICLR 2024) · better on 8 of 8 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
- SiReRAG: Indexing Similar and Related Information for Multihop Reasoning (ICLR 2025 · similarity 0.90)
- A Theory for Token-Level Harmonization in Retrieval-Augmented Generation (ICLR 2025 · similarity 0.87)
- A Systematic Exploration of Knowledge Graph Alignment with Large Language Models in Retrieval Augmented Generation (AAAI 2025 · similarity 0.86)
- StructRAG: Boosting Knowledge Intensive Reasoning of LLMs via Inference-time Hybrid Information Structurization (ICLR 2025 · similarity 0.86)
- Can Knowledge-Graph-based Retrieval Augmented Generation Really Retrieve What You Need? (NeurIPS 2025 · similarity 0.86)
- RAPTOR: Recursive Abstractive Processing for Tree-Organized Retrieval (ICLR 2024 · similarity 0.86)
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.
