mlsanityData to 14 Oct 2026

SARA: Selective and Adaptive Retrieval-augmented Generation with Context Compression

arXiv 2507.05633 ↗First posted 8 Jul 2025cs.CL cs.AI cs.IRcited by 10

On the map: Retrieval-augmented generation → Generation and context integration

also in 1 more
Abstract (arXiv)

Retrieval-augmented Generation (RAG) extends large language models (LLMs) with external knowledge but faces key challenges: restricted effective context length and redundancy in retrieved documents. Pure compression-based approaches reduce input size but often discard fine-grained details essential for factual accuracy. We propose SARA, a unified RAG framework that balances local precision and global knowledge coverage under tight context budgets. SARA combines natural-language text snippets with semantic compression vectors to jointly enhance context efficiency and answer correctness. It represents contexts at two complementary levels: 1) fine-grained natural-language spans that preserve critical entities and numerical values, and 2) compact, interpretable vectors that summarize high-level semantics. An iterative evidence-selection module employs the compression vectors for dynamic reranking of contexts. Across 9 datasets and 5 open-source LLMs spanning 3 model families (Mistral, Llama, and Gemma), SARA consistently improves answer relevance (+17.71), answer correctness (+13.72), and semantic similarity (+15.53), demonstrating the importance of integrating textual and compressed representations for robust, context-efficient RAG.

Key facts. Cited by 10 papers in the map (data through 14 Oct 2026). Placed in Generation and context integration, one of 15 directions in Retrieval-augmented generation. Builds on 56 of its 78 references that are in the map. First posted on arXiv 8 Jul 2025.

Builds on

56 of 78 references.

Closest ancestors · 11

Same direction, newest first

  1. 2025Provence: efficient and robust context pruning for retrieval-augmented generation27 Jan
  2. 2025PISCO: Pretty Simple Compression for Retrieval-Augmented Generation27 Jan
  3. 2024Jasper and Stella: distillation of SOTA embedding models26 Dec
  4. 2024EXIT: Context-Aware Extractive Compression for Enhancing Retrieval-Augmented Generation17 Dec
  5. 2024AdaComp: Extractive Context Compression with Adaptive Predictor for Retrieval-Augmented Large Language Models3 Sep
  6. 2024CompAct: Compressing Retrieved Documents Actively for Question Answering12 Jul
  7. 2024RankRAG: Unifying Context Ranking with Retrieval-Augmented Generation in LLMs2 Jul
  8. 2024xRAG: Extreme Context Compression for Retrieval-augmented Generation with One Token22 May
  9. 2023RECOMP: Improving Retrieval-Augmented LMs with Compression and Selective Augmentation6 Oct
  10. 2022Atlas: Few-shot Learning with Retrieval Augmented Language Models5 Aug
  11. 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks22 May

Other references · 45

  1. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,305
  2. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,380
  3. 2019Sentence-BERT: Sentence Embeddings using Siamese BERT-Networkscited by 9,170
  4. 2023Visual Instruction Tuningcited by 7,869
  5. 2018HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answeringcited by 3,754
  6. 2023Mistral 7Bcited by 3,164
  7. 2017TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehensioncited by 2,955
  8. 2023Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflectioncited by 1,839
  9. 2018Know What You Don't Know: Unanswerable Questions for SQuADcited by 1,554
  10. 2024From Local to Global: A Graph RAG Approach to Query-Focused Summarizationcited by 1,453
  11. 2023LongBench: A Bilingual, Multitask Benchmark for Long Context Understandingcited by 1,417
  12. 2020Constructing A Multi-hop QA Dataset for Comprehensive Evaluation of Reasoning Stepscited by 1,282
33 more
  1. 2021BEIR: A Heterogenous Benchmark for Zero-shot Evaluation of Information Retrieval Modelscited by 1,252
  2. 2020MPNet: Masked and Permuted Pre-training for Language Understandingcited by 981
  3. 2023Towards General Text Embeddings with Multi-stage Contrastive Learningcited by 723
  4. 2022MTEB: Massive Text Embedding Benchmarkcited by 640
  5. 2017The NarrativeQA Reading Comprehension Challengecited by 597
  6. 2023Enabling Large Language Models to Generate Text with Citationscited by 595
  7. 2024RAPTOR: Recursive Abstractive Processing for Tree-Organized Retrievalcited by 532
  8. 2021A Dataset of Information-Seeking Questions and Answers Anchored in Research Paperscited by 486
  9. 2023LongLLMLingua: Accelerating and Enhancing LLMs in Long Context Scenarios via Prompt Compressioncited by 453
  10. 2023Ragas: Automated Evaluation of Retrieval Augmented Generationcited by 418
  11. 2023Learning to Compress Prompts with Gist Tokenscited by 390
  12. 2021QMSum: A New Benchmark for Query-based Multi-domain Meeting Summarizationcited by 365
  13. 2023Adapting Language Models to Compress Contextscited by 356
  14. 2024LLMLingua-2: Data Distillation for Efficient and Faithful Task-Agnostic Prompt Compressioncited by 348
  15. 2023LLMLingua: Compressing Prompts for Accelerated Inference of Large Language Modelscited by 287
  16. 2021QuALITY: Question Answering with Long Input Texts, Yes!cited by 208
  17. 2023LongNet: Scaling Transformers to 1,000,000,000 Tokenscited by 204
  18. 2024Leave No Context Behind: Efficient Infinite Context Transformers with Infini-attentioncited by 200
  19. 2023In-context Autoencoder for Context Compression in a Large Language Modelcited by 199
  20. 2024LLM Maybe LongLM: Self-Extend LLM Context Window Without Tuningcited by 136
  21. 2024InstructRAG: Instructing Retrieval-Augmented Generation via Self-Synthesized Rationalescited by 101
  22. 2024Searching for Best Practices in Retrieval-Augmented Generationcited by 83
  23. 2023Better to Ask in English: Cross-Lingual Evaluation of Large Language Models for Healthcare Queriescited by 82
  24. 2024Found in the Middle: How Language Models Use Long Contexts Better via Plug-and-Play Positional Encodingcited by 72
  25. 2021Semantic Answer Similarity for Evaluating Question Answering Modelscited by 47
  26. 2023Compressed Context Memory For Online Language Model Interactioncited by 43
  27. 2024GraphReader: Building Graph-based Agent to Enhance Long-Context Abilities of Large Language Modelscited by 43
  28. 2025CultureVLM: Characterizing and Improving Cultural Understanding of Vision-Language Models for over 100 Countriescited by 42
  29. 2024ProteinGPT: Multimodal LLM for Protein Property Prediction and Structure Understandingcited by 23
  30. 2024OG-RAG: Ontology-Grounded Retrieval-Augmented Generation For Large Language Modelscited by 23
  31. 2024Understanding Retrieval Robustness for Retrieval-Augmented Image Captioningcited by 9
  32. 2024Semi-supervised Fine-tuning for Large Language Modelscited by 8
  33. 2025GraPPI: A Retrieve-Divide-Solve GraphRAG Framework for Large-scale Protein-protein Interaction Explorationcited by 3

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 2; 2026: 8.

10 papers.

  1. 2026TextReg: Mitigating Prompt Distributional Overfitting via Regularized Text-Space Optimizationcited by 3
  2. 2026ZipRL: Adaptive Multi-Turn Context Compression with Hindsight Response Replaycited by 3
  3. 2025When More Retrieval Hurts: Retrieval-Augmented Code Review Generationcited by 2
  4. 2025Concept than Document: Context Compression via AMR-based Conceptual Entropycited by 2
  5. 2026CompactRAG: Reducing LLM Calls and Token Overhead in Multi-Hop Question Answeringcited by 1
  6. 2026Understand Then Memory: A Cognitive Gist-Driven RAG Framework with Global Semantic Diffusioncited by 0
  7. 2026Rethinking Retrieval-Augmentation as Synthesis: A Query-Aware Context Merging Approachcited by 0
  8. 2026When Context Misleads: Intent-Guided Decoding for Robust Retrieval-Augmented Generationcited by 0
  9. 2026Compression Beyond the Uncompressed: A Two-Stage Training Recipe for Soft Context Compression in RAGcited by 0
  10. 2026REVA: Reusable Evidence View Aggregation for Context-Efficient RAG Servingcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024Contextual Compression in Retrieval-Augmented Generation for Large Language Models: A Survey
  2. 2025Enhancing RAG Efficiency with Adaptive Context Compression
  3. 2025AttnComp: Attention-Guided Adaptive Context Compression for Retrieval-Augmented Generation
  4. 2025Grounding Long-Context Reasoning with Contextual Normalization for Retrieval-Augmented Generation
  5. 2025Less Is More: Elevating RAG via Performance-Driven Context Compression
  6. 2025On the Influence of Context Size and Model Choice in Retrieval-Augmented Generation Systems
  7. 2025Retrieval-Augmented Generation: A Comprehensive Survey of Architectures, Enhancements, and Robustness Frontiers
  8. 2025Enhancing Retrieval-Augmented Generation: A Study of Best Practices