mlsanityData to 14 Oct 2026

When More Documents Hurt RAG: Mitigating Vector Search Dilution with Domain-Scoped, Model-Agnostic Retrieval

arXiv 2606.11350 ↗First posted 9 Jun 2026cs.CL cs.IRcited by 0

On the map: Retrieval-augmented generation → RAG pipeline design and evaluation

Abstract (arXiv)

Retrieval-augmented generation (RAG) degrades when applied to large, heterogeneous document collections. In RAG systems, dense similarity becomes less discriminative, causing top-k retrieval to return semantically similar but contextually incorrect chunks. We refer to this failure mode as vector search dilution. Even under hybrid dense+sparse retrieval, we first observed it in a deployed Wyoming Department of Transportation (WYDOT) system; in a controlled experiment, growing the corpus from 54 to 1,128 documents cut precision@10 by 63% and recall of the answer-bearing document by 23%. Similar patterns were observed in six other DOTs. To diagnose this issue, we introduce MASDR-RAG (Metadata-Scoped Domain Retrieval for RAG) as a controlled testbed over 200 human-validated queries, five LLM backbones, six public benchmark corpora, and two index stacks. In MASDR-RAG, domain scoping is the load-bearing component. We found hard partitioning to be unsafe, whereas training-free soft scoping is safe when the scope structure is real. Furthermore, we can discover these scopes without hand-curated metadata. However, multi-agent orchestration triggers a precision-faithfulness paradox that survives format controls, prompt controls, and blind human evaluation. Based on these varied outcomes, our practical recommendation is simple: soft-scope first, then perform a single synthesis call.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in RAG pipeline design and evaluation, one of 15 directions in Retrieval-augmented generation. Builds on 26 of its 39 references that are in the map. First posted on arXiv 9 Jun 2026.

Builds on

26 of 39 references.

Closest ancestors · 4

Same direction, newest first

  1. 2024Long-Context LLMs Meet RAG: Overcoming Challenges for Long Inputs in RAG8 Oct
  2. 2024MultiHop-RAG: Benchmarking Retrieval-Augmented Generation for Multi-Hop Queries27 Jan
  3. 2024The Power of Noise: Redefining Retrieval for RAG Systems26 Jan
  4. 2024Seven Failure Points When Engineering a Retrieval Augmented Generation System11 Jan

Other references · 22

  1. 2024The Llama 3 Herd of Modelscited by 14,715
  2. 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,297
  3. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,695
  4. 2023Toolformer: Language Models Can Teach Themselves to Use Toolscited by 4,182
  5. 2020Dense Passage Retrieval for Open-Domain Question Answeringcited by 3,824
  6. 2018HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answeringcited by 3,754
  7. 2017Billion-scale similarity search with GPUscited by 2,497
  8. 2020REALM: Retrieval-Augmented Language Model Pre-Trainingcited by 2,141
  9. 2023Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflectioncited by 1,839
  10. 2024From Local to Global: A Graph RAG Approach to Query-Focused Summarizationcited by 1,453
  11. 2020ColBERT: Efficient and Effective Passage Search via Contextualized Late Interaction over BERTcited by 1,178
  12. 2022Measuring and Narrowing the Compositionality Gap in Language Modelscited by 1,075
10 more
  1. 2022Interleaving Retrieval with Chain-of-Thought Reasoning for Knowledge-Intensive Multi-Step Questionscited by 923
  2. 2016Efficient and robust approximate nearest neighbor search using Hierarchical Navigable Small World graphscited by 861
  3. 2023Large Language Models Can Be Easily Distracted by Irrelevant Contextcited by 841
  4. 2021ColBERTv2: Effective and Efficient Retrieval via Lightweight Late Interactioncited by 476
  5. 2024Adaptive-RAG: Learning to Adapt Retrieval-Augmented Large Language Models through Question Complexitycited by 451
  6. 2020Document Ranking with a Pretrained Sequence-to-Sequence Modelcited by 376
  7. 2024HELMET: How to Evaluate Long-Context Language Models Effectively and Thoroughlycited by 125
  8. 2022From Distillation to Hard Negative Sampling: Making Sparse Neural IR Models More Effectivecited by 98
  9. 2020The Curse of Dense Low-Dimensional Information Retrieval for Large Index Sizescited by 48
  10. 2026SCOUT-RAG: Scalable and Cost-Efficient Unifying Traversal for Agentic Graph-RAG over Distributed Domainscited by 2

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026BM25 Wins at Scale: A Scaling Study of Retrieval-Augmented Generation Paradigms
  2. 2026RIT-RAG: Navigating Document Corpora with Retrieval-Induced Trees
  3. 2025MODE: Mixture of Document Experts for RAG
  4. 2026A-RAG: Scaling Agentic Retrieval-Augmented Generation via Hierarchical Retrieval Interfaces
  5. 2026CHOP: Chunkwise Context-Preserving Framework for RAG on Multi Documents
  6. 2026Mapping the RAG Landscape: A Four Axis Taxonomy of Efficiency, Defense, Interactivity, and Reasoning
  7. 2025Less LLM, More Documents: Searching for Improved RAG
  8. 2026Agent-Orchestrated Adaptive RAG: A Comparative Study on Structured and Multi-Hop Retrieval