mlsanityData to 14 Oct 2026

PRIME: Pre-Retrieval Index Modeling Enhancement for Structured Document Retrieval-Real

arXiv 2610.13296 ↗First posted 9 Oct 2026cs.IR cs.AIcited by 0

New in LLM agents and tool use; no direction yet. LLM agents and tool use

Abstract (arXiv)

High-precision retrieval over external knowledge bases is essential for LLM agents to operate effectively. However, existing Information Retrieval (IR) research primarily optimizes query-side retrievers (e.g., ColBERT, BGE) and generators (e.g., Self-RAG, GraphRAG), while reducing corpus-side representation to chunking. This leaves index-side quality an underexplored lever. We propose Pre-Retrieval Index Modeling & Enhancement (PRIME), an index-side optimization framework built on two complementary tracks: an LLM-driven Semantic Restructuring (SR) track that transforms long documents into structured, information-dense sections, and a deterministic Metadata Synthesis (MS) and Hierarchical Retrieval (HR) track that extracts semantic anchors and heading structure for coarse-to-fine section retrieval. Our evaluations show that deterministic metadata matches structure-aware chunking's retrieval quality while cutting embedding computation by 45.8 × and index vector count by 22.4 ×. On top of these gains, Semantic Restructuring (SR) further improves Answer-F1 by 13.7% in full deployment, with characterized quality-fidelity-cost trade-offs. Across QASPER, BRIGHT, and NFCorpus, we further characterize three qualitative regimes that determine whether index-side optimization helps or degrades retrieval, while showing that these index-side effects are neutralized under a strong reranker.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 25 of its 35 references that are in the map. First posted on arXiv 9 Oct 2026.

Builds on

25 of 35 references.

References · 25

  1. 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,297
  2. 2019Sentence-BERT: Sentence Embeddings using Siamese BERT-Networkscited by 9,170
  3. 2020Dense Passage Retrieval for Open-Domain Question Answeringcited by 3,824
  4. 2023Lost in the Middle: How Language Models Use Long Contextscited by 3,455
  5. 2020REALM: Retrieval-Augmented Language Model Pre-Trainingcited by 2,141
  6. 2024From Local to Global: A Graph RAG Approach to Query-Focused Summarizationcited by 1,453
  7. 2021Improving language models by retrieving from trillions of tokenscited by 1,293
  8. 2021BEIR: A Heterogenous Benchmark for Zero-shot Evaluation of Information Retrieval Modelscited by 1,252
  9. 2021Unsupervised Dense Information Retrieval with Contrastive Learningcited by 1,216
  10. 2020ColBERT: Efficient and Effective Passage Search via Contextualized Late Interaction over BERTcited by 1,178
  11. 2022Atlas: Few-shot Learning with Retrieval Augmented Language Modelscited by 952
  12. 2016Efficient and robust approximate nearest neighbor search using Hierarchical Navigable Small World graphscited by 861
13 more
  1. 2020Approximate Nearest Neighbor Negative Contrastive Learning for Dense Text Retrievalcited by 849
  2. 2019Passage Re-ranking with BERTcited by 645
  3. 2023C-Pack: Packed Resources For General Chinese Embeddingscited by 578
  4. 2024RAPTOR: Recursive Abstractive Processing for Tree-Organized Retrievalcited by 532
  5. 2022Precise Zero-Shot Dense Retrieval without Relevance Labelscited by 517
  6. 2021A Dataset of Information-Seeking Questions and Answers Anchored in Research Paperscited by 486
  7. 2021SPLADE: Sparse Lexical and Expansion Model for First Stage Rankingcited by 286
  8. 2019Document Expansion by Query Predictioncited by 231
  9. 2024BRIGHT: A Realistic and Challenging Benchmark for Reasoning-Intensive Retrievalcited by 158
  10. 2025PathRAG: Pruning Graph-based Retrieval Augmented Generation with Relational Pathscited by 61
  11. 2025LeanRAG: Knowledge-Graph-Based Generation with Semantic Aggregation and Hierarchical Retrievalcited by 16
  12. 2024Multi-view Content-aware Indexing for Long Document Retrievalcited by 5
  13. 2026Beyond the Reranker: Do RAG Retrieval Enhancements Help Once a Strong Reranker Is Present?cited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Structure-Aware Semantic Chunking with Title-Chain Prefixes: A 1600-Query Evaluation and the Measurement Trap in Text-Transform Ablations
  2. 2026Does Document Structure Help Dense Retrieval? A Placebo-Controlled Ablation of Four Mechanisms Across Two Corpora
  3. 2026CHOP: Chunkwise Context-Preserving Framework for RAG on Multi Documents
  4. 2026Evaluating Chunking Strategies For Retrieval-Augmented Generation in Oil and Gas Enterprise Documents
  5. 2026A Systematic Investigation of Document Chunking Strategies and Embedding Sensitivity
  6. 2026Query Expansion Should Be Coordinated: Dense Expands, Sparse Anchors
  7. 2026When More Documents Hurt RAG: Mitigating Vector Search Dilution with Domain-Scoped, Model-Agnostic Retrieval
  8. 2025REAL-MM-RAG: A Real-World Multi-Modal Retrieval Benchmark