mlsanityData to 14 Oct 2026

Principled and Scalable Diversity-Aware Retrieval via Cardinality-Constrained Binary Quadratic Programming

arXiv 2604.02554 ↗First posted 2 Apr 2026cs.CL cs.IRcited by 2

On the map: Retrieval-augmented generation → RAG robustness and adaptive retrieval

Abstract (arXiv)

Diversity-aware retrieval is essential for Retrieval-Augmented Generation (RAG), yet existing methods lack theoretical guarantees and face scalability issues as the number of retrieved passages k increases. We propose a principled formulation of diversity retrieval as a cardinality-constrained binary quadratic programming (CCBQP), which explicitly balances relevance and semantic diversity through an interpretable trade-off parameter. Inspired by recent advances in combinatorial optimization, we develop a non-convex tight continuous relaxation and a Frank--Wolfe based algorithm with landscape analysis and convergence guarantees. Extensive experiments demonstrate that our method consistently dominates baselines on the relevance-diversity Pareto frontier, while achieving significant speedup.

Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Placed in RAG robustness and adaptive retrieval, one of 15 directions in Retrieval-augmented generation. Builds on 15 of its 36 references that are in the map. First posted on arXiv 2 Apr 2026.

Builds on

15 of 36 references.

Closest ancestors · 1

Same direction, newest first

  1. 2026DF-RAG: Query-Aware Diversity for Retrieval-Augmented Generation23 Jan

Other references · 14

  1. 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,297
  2. 2023Efficient Memory Management for Large Language Model Serving with PagedAttentioncited by 5,944
  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. 2024M3-Embedding: Multi-Linguality, Multi-Functionality, Multi-Granularity Text Embeddings Through Self-Knowledge Distillationcited by 1,073
  6. 2019How Contextual are Contextualized Word Representations? Comparing the Geometry of BERT, ELMo, and GPT-2 Embeddingscited by 784
  7. 2022ASQA: Factoid Questions Meet Long-Form Answerscited by 234
  8. 2016Convergence Rate of Frank-Wolfe for Non-Convex Objectivescited by 63
  9. 2022QAMPARI: An Open-domain Question Answering Benchmark for Questions with Many Answers from Multiple Paragraphscited by 42
  10. 2022Conditional Gradient Methodscited by 17
  11. 2024Vector Retrieval with Similarity and Diversity: How Hard Is It?cited by 3
  12. 2013The Sparse Principal Component of a Constant-rank Matrixcited by 2
2 more
  1. 2025MUSS: Multilevel Subset Selection for Relevance and Diversitycited by 2
  2. 2024On Densest k-Subgraph Mining and Diagonal Loading: Optimization Landscape and Finite-Step Exact Convergence Analysiscited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 2.

2 papers.

  1. 2026WARP: Wasserstein-Aligned RAG for Population Opinionscited by 0
  2. 2026Finding the Right Balance: Relevance and Diversity in LLM Retrievalcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026DIVERGE: Diversity-Enhanced RAG for Open-Ended Information Seeking
  2. 2026Fairness-Aware Retrieval Optimization for Retrieval-Augmented Generation
  3. 2026Welfarist Formulations for Diverse Similarity Search
  4. 2025Vendi-RAG: Adaptively Trading-Off Diversity And Quality Significantly Improves Retrieval Augmented Generation With LLMs
  5. 2026Scaling DPPs for RAG: Density Meets Diversity
  6. 2026ISO-RAG: Isoperimetric Noise Control for Retrieval-Augmented Generation
  7. 2026Tail-Aware Adaptive-k: Query-Adaptive Context Selection for Retrieval-Augmented Generation
  8. 2024Diversify-verify-adapt: Efficient and Robust Retrieval-Augmented Ambiguous Question Answering