NuggetIndex: Governed Atomic Retrieval for Maintainable RAG
sigir26-163 · Full research paper · Saber Zerhoudi, Michael Granitzer, Jelena Mitrović
Learned Sparse Retrieval Efficiency · 50% Evidence Selection and Context Refin · 25% Knowledge Conflict and Faithfulness · 25%
Abstract
Retrieval-augmented generation (RAG) systems are frequently evaluated via fact-based metrics, yet standard implementations retrieve passages or static propositions. This unit mismatch between evaluation and retrieval objects hinders maintenance when corpora evolve and fails to capture superseded facts or source disagreements. We propose NuggetIndex, a retrieval system that stores atomic information units as managed records, so called nuggets. Each record maintains links to evidence, a temporal validity interval, and a lifecycle state. By filtering invalid or deprecated nuggets prior to ranking, the system prevents the inclusion of outdated information. We evaluate the approach using a nuggetized MS MARCO subset, a temporal Wikipedia QA dataset, and a multi-hop QA task. Against passage and unmanaged proposition retrieval baselines, NuggetIndex improves nugget recall by 42%, increases temporal correctness by 9 percentage points without the recall collapse observed in time-filtered baselines, and reduces conflict rates by 55%. The compact nugget format reduces generator input length by 64% while enabling lightweight index structures suitable for browser-based and resource-constrained deployment. We release our implementation, datasets, and evaluation scripts
Assigned reviewers
| # | Reviewer | Match | Why (closest evidence publications in the taxonomy) | |
|---|---|---|---|---|
| 1 | Huiqiang Jiang | 69 | 63 SCBench: A KV Cache-Centric Analysis of Long-Context Methods (2025) Learned Sparse Retrieval Efficiency 63 LLMLingua: Compressing Prompts for Accelerated Inference of Large Language Models (2023) Learned Sparse Retrieval Efficiency 63 LongLLMLingua: Accelerating and Enhancing LLMs in Long Context Scenarios via Prompt Compression (2024) Learned Sparse Retrieval Efficiency |
|
| 2 | Jing Liu 0022 | 53 | 63 Reinforced Informativeness Optimization for Long-Form Retrieval-Augmented Generation (2026) Evidence Selection and Context Refinement 51 Investigating the Factual Knowledge Boundary of Large Language Models with Retrieval Augmentation (2025) Knowledge Conflict and Faithfulness 51 REAR: A Relevance-Aware Retrieval-Augmented Framework for Open-Domain Question Answering (2024) Evidence Selection and Context Refinement |
Area chair
| Area chair | Match | |
|---|---|---|
| Xiaoxi Li 0005 | 48 |
Area-chair candidates (5)
| Xiaoxi Li 0005 assigned | 48 | 15 papers | |
| Hongjin Qian | 45 | 12 papers | |
| Xiaoxue Zang | 45 | 4 papers | |
| Jun Xu 0001 | 43 | 2 papers | |
| Fernando Diaz 0001 | 40 | 12 papers |
Candidates from the panel ranked by match score
| # | Reviewer | Match | Load | Why | |
|---|---|---|---|---|---|
| 1 | Huiqiang Jiang assigned | 69 | 3 | 63 SCBench: A KV Cache-Centric Analysis of Long-Context Methods (2025) Learned Sparse Retrieval Efficiency 63 LLMLingua: Compressing Prompts for Accelerated Inference of Large Language Models (2023) Learned Sparse Retrieval Efficiency |
|
| 2 | Franco Maria Nardini | 63 | 3 | 65 Efficient Re-ranking with Cross-encoders via Early Exit (2025) Learned Sparse Retrieval Efficiency 65 Quality versus efficiency in document scoring with learning-to-rank models (2016) Learned Sparse Retrieval Efficiency |
|
| 3 | Antonio Mallia | 61 | 3 | 65 E2Rank: Efficient and Effective Layer-Wise Reranking (2025) Learned Sparse Retrieval Efficiency 63 Faster Learned Sparse Retrieval with Block-Max Pruning (2024) Learned Sparse Retrieval Efficiency |
|
| 4 | Sebastian Bruch 0001 | 55 | 3 | 63 Forward Index Compression for Learned Sparse Retrieval (2026) Learned Sparse Retrieval Efficiency 63 Efficient Inverted Indexes for Approximate Retrieval over Learned Sparse Representations (2024) Learned Sparse Retrieval Efficiency |
|
| 5 | Jing Liu 0022 assigned | 53 | 3 | 63 Reinforced Informativeness Optimization for Long-Form Retrieval-Augmented Generation (2026) Evidence Selection and Context Refinement 51 Investigating the Factual Knowledge Boundary of Large Language Models with Retrieval Augmentation (2025) Knowledge Conflict and Faithfulness |
|
| 6 | Xueguang Ma | 53 | 3 | 65 Gosling Grows Up: Retrieval with Learned Dense and Sparse Representations Using Anserini (2025) Learned Sparse Retrieval Efficiency 64 Pyserini: A Python Toolkit for Reproducible Information Retrieval Research with Sparse and Dense Representations (2021) Learned Sparse Retrieval Efficiency |
|
| 7 | Sheng-Chieh Lin | 51 | 3 | 66 One Blade for One Purpose: Advancing Math Information Retrieval using Hybrid Search (2023) Learned Sparse Retrieval Efficiency 65 Pyserini: A Python Toolkit for Reproducible Information Retrieval Research with Sparse and Dense Representations (2021) Learned Sparse Retrieval Efficiency |
|
| 8 | Xinyu Ma 0001 | 50 | 3 | 62 Probe-and-Fetch: Dynamic KV Cache Pruning for Accelerated Long-Context Inference in Web-Scale AI Search (2026) Efficient and Long-Context RAG 51 Direct Retrieval-augmented Optimization: Synergizing Knowledge Selection and Language Models (2026) Evidence Selection and Context Refinement |
|
| 10 | Xiao Wang 0043 | 47 | 3 | 65 An Inspection of the Reproducibility and Replicability of TCT-ColBERT (2022) Learned Sparse Retrieval Efficiency 65 Reproducibility, Replicability, and Insights into Dense Multi-Representation Retrieval Models: from ColBERT to Col (2023) Learned Sparse Retrieval Efficiency |
|
| 11 | Daiting Shi | 47 | 3 | 64 Pre-trained Language Model-based Retrieval and Ranking for Web Search (2023) Learned Sparse Retrieval Efficiency 62 Pre-trained Language Model based Ranking in Baidu Search (2021) Learned Sparse Retrieval Efficiency |
|
| 12 | Jurek Leonhardt | 47 | 3 | 64 Efficient Neural Ranking using Forward Indexes (2022) Learned Sparse Retrieval Efficiency 64 Efficient Neural Ranking Using Forward Indexes and Lightweight Encoders (2024) Learned Sparse Retrieval Efficiency |
|
| 13 | Shu Zhou 0002 | 46 | 3 | 62 Activation Caching for Retrieval-Augmented Generation (2026) Efficient and Long-Context RAG 46 Calibrating Uncertainty with Cross-Model Consistency for LLM Hallucination Mitigation (2026) Knowledge Conflict and Faithfulness |
|
| 16 | Jinheon Baek | 45 | 3 | 81 Efficient Long Context Language Model Retrieval with Compression (2025) Learned Sparse Retrieval Efficiency 50 Knowledge-Augmented Reasoning Distillation for Small Language Models in Knowledge-Intensive Tasks (2023) Retrieval-Augmented and Open-Domain QA |
|
| 17 | Yujia Zhou 0002 | 45 | 3 | 55 Grounding Language Model with Chunking-Free In-Context Retrieval (2024) Generative Retrieval 55 CorpusLM: Towards a Unified Language Model on Corpus for Knowledge-Intensive Tasks (2024) Generative Retrieval |
|
| 18 | Weiwei Sun 0001 | 44 | 3 | 51 Direct Retrieval-augmented Optimization: Synergizing Knowledge Selection and Language Models (2026) Evidence Selection and Context Refinement 50 Contrastive Learning Reduces Hallucination in Conversations (2023) Knowledge Conflict and Faithfulness |
|
| 19 | Yu-An Liu 0028 | 44 | 3 | 50 AdversarialCoT: Single-Document Retrieval Poisoning for LLM Reasoning (2026) Knowledge Conflict and Faithfulness 50 Compete to Complete: Co-opetition Adversarial Learning for Retrieval-Augmented Generation (2026) Knowledge Conflict and Faithfulness |
|
| 20 | Yuchen Li 0006 | 44 | 3 | 62 Probe-and-Fetch: Dynamic KV Cache Pruning for Accelerated Long-Context Inference in Web-Scale AI Search (2026) Efficient and Long-Context RAG 51 Enhancing Retrieval-Augmented Generation via Evidence Tree Search (2025) Evidence Selection and Context Refinement |
|
| 21 | Jialong Wu 0007 | 44 | 3 | 63 SCOPE: Optimizing Key-Value Cache Compression in Long-context Generation (2025) Learned Sparse Retrieval Efficiency 46 WebWalker: Benchmarking LLMs in Web Traversal (2025) Agentic and Deep Search RAG |
|
| 22 | Di Wu 0054 | 44 | 3 | 52 LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory (2025) Conversational Search and Retrieval 50 Synchronous Faithfulness Monitoring for Trustworthy Retrieval-Augmented Generation (2024) Knowledge Conflict and Faithfulness |
|
| 23 | Navve Wasserman | 44 | 3 | 52 DocReRank: Single-Page Hard Negative Query Generation for Training Multi-Modal RAG Rerankers (2025) Universal Multimodal Embedding and MLLM Retrieval 52 REAL-MM-RAG: A Real-World Multi-Modal Retrieval Benchmark (2025) Universal Multimodal Embedding and MLLM Retrieval |
|
| 25 | Venktesh V | 43 | 3 | 65 Breaking the Lens of the Telescope: Online Relevance Estimation over Large Retrieval Sets (2025) Learned Sparse Retrieval Efficiency 57 The Surprising Effectiveness of Rankers trained on Expanded Queries (2024) Evidence Selection and Context Refinement |
|
| 26 | Jingtao Zhan | 43 | 3 | 63 Joint Optimization of Multi-vector Representation with Product Quantization (2022) Learned Sparse Retrieval Efficiency 43 Combining Multiple Supervision for Robust Zero-Shot Dense Retrieval (2024) Sparse and Dense Retrieval Encoders |
|
| 27 | Shengyao Zhuang | 43 | 3 | 65 Layer-wise Token Compression for Efficient Document Reranking (2026) Learned Sparse Retrieval Efficiency 64 LACONIC: Dense-Level Effectiveness for Scalable Sparse Retrieval via a Two-Phase Training Curriculum (2026) Learned Sparse Retrieval Efficiency |
|
| 28 | Shuaiqiang Wang | 43 | 3 | 51 Enhancing Retrieval-Augmented Generation via Evidence Tree Search (2025) Evidence Selection and Context Refinement 37 Is ChatGPT Good at Search? Investigating Large Language Models as Re-Ranking Agents (2023) Text Embedding and Representation Learning |
|
| 29 | Ruqing Zhang 0001 | 43 | 3 | 55 Boosting Retrieval-Augmented Generation with Generation-Augmented Retrieval: A Co-Training Approach (2025) Generative Retrieval 41 B-PROP: Bootstrapped Pre-training with Representative Words Prediction for Ad-hoc Retrieval (2021) Sparse and Dense Retrieval Encoders |
|
| 30 | Corby Rosset | 42 | 3 | 65 Dodo: Dynamic Contextual Compression for Decoder-only LMs (2024) Learned Sparse Retrieval Efficiency 63 Optimizing Query Evaluations Using Reinforcement Learning for Web Search (2018) Efficient Retrieval and Indexing |
|
| 31 | Xiao Zhang 0034 | 42 | 3 | 50 ReDeEP: Detecting Hallucination in Retrieval-Augmented Generation via Mechanistic Interpretability (2025) Knowledge Conflict and Faithfulness 50 LargePiG for Hallucination-Free Query Generation: Your Large Language Model is Secretly a Pointer Generator (2025) Knowledge Conflict and Faithfulness |
|
| 32 | Keping Bi | 42 | 3 | 53 Beyond Relevance: Utility-Centric Retrieval in the LLM Era (2026) Evidence Selection and Context Refinement 51 Utility-Focused LLM Annotation for Retrieval and Retrieval-Augmented Generation (2025) Evidence Selection and Context Refinement |
|
| 33 | Weipeng Chen | 42 | 3 | 65 PQCache: Product Quantization-based KVCache for Long Context LLM Inference (2025) Learned Sparse Retrieval Efficiency 56 RichRAG: Crafting Rich Responses for Multi-faceted Queries in Retrieval-Augmented Generation (2025) Evidence Selection and Context Refinement |
|
| 34 | Negar Arabzadeh | 41 | 3 | 58 Can QPP Choose the Right Query variant? Evaluating Query Variant Selection for RAG Pipelines (2026) Evidence Selection and Context Refinement 50 Predicting Efficiency/Effectiveness Trade-offs for Dense vs. Sparse Retrieval Strategy Selection (2021) Efficient Retrieval and Indexing |