Dense Passage Retrieval for Open-Domain Question Answering

2020/04/10 by Karpukhin, Vladimir, Oğuz, Barlas, Min, Sewon +5 · 485 citations
#Computation and Language (cs.CL) #FOS: Computer and information sciences

paper · doi:10.48550/arxiv.2004.04906

Abstract

Open-domain question answering relies on efficient passage retrieval to select candidate contexts, where traditional sparse vector space models, such as TF-IDF or BM25, are the de facto method. In this work, we show that retrieval can be practically implemented using dense representations alone, where embeddings are learned from a small number of questions and passages by a simple dual-encoder framework. When evaluated on a wide range of open-domain QA datasets, our dense retriever outperforms a strong Lucene-BM25 system largely by 9%-19% absolute in terms of top-20 passage retrieval accuracy, and helps our end-to-end QA system establish new state-of-the-art on multiple open-domain QA benchmarks.

Cited by

• DeCoRAG: Cognitive Decoupling and Semantic-Aware Cropping for Complex Document Understanding
• From transcription to semantic corpus analysis: unsupervised learning of sentence representations for ancient languages
• Keep It InMind: Benchmarking the Implicit-Association Blind Spot in Agent Memory
• From Unstructured Recall to Schema-Grounded Memory: Reliable AI Memory via Iterative, Schema-Aware Extraction
• MemChain: Learning Interpretable Memory Traces for Memory-Augmented LLM Agents
• Isolated but Exposed: Persistence-Based Memory Extraction Attack on LLM Agents
• TriShieldRAG: A Three-Ring Defense-in-Depth Framework Against Knowledge Corruption in Retrieval-Augmented Generation
• Offline-to-Online Creative Optimization with Generative Models and Adaptive Testing
• A New Role for Relevance: Guiding Corpus Interaction in Agentic Search
• KAP: Bridging the Knowledge Selection-Runtime Consumption Gap in LLM Systems
• Do Current Retrievers Cover All the Evidence? A Controlled Study of Conjunctive Cross-Page Retrieval
• VecTree-RAG: An Agentic Retrieval-Augmented Generation Framework Combining Vector and Tree Retrieval for Efficiency and Accuracy
• LazyMem: Retrieve Broadly, Construct Selectively for Efficient Long-Term Agent Memory
• Salient Knowledge Pathways: Sparse Cross-Modal Routing for Efficient Knowledge-Intensive Multimodal Question Answering
• Bekko Embedding: Parameter-Efficient Multilingual Retrieval with Ultra-Compact Encoders
• The Case Against Generation for Retrieval: Discriminative Language Models as Effective Retrievers
• HVM-GraphRAG: Holistic-View Multimodal Graph Retrieval-Augmented Generation on Complex Document
• FinAbstain: Uncertainty-Calibrated Multimodal RAG for Selective Financial Forecasting
• ScalableRAG: High-Quality RAG at Zero Ingestion Cost
• Aethel: A Reproducible Graph-Retrieval Framework for Multi-Hop Financial Diligence
• SciClaimSeekers at CheckThat! 2026: Retrieving Scientific Sources for Social Media Claims with LLM Reranking
• When Thinking Before Retrieval Hurts: TraceBound Diagnostics for Adaptive Knowledge-Graph Retrieval
• MetaSyn: A Benchmark for LLM Agents on Meta-Analysis Articles from Nature Portfolio
• AI-Assisted Knowledge Access for Legacy Enterprise Asset Management in Energy Operations: A Practical Retrieval System
• TRACE: Business Rule-Grounded Reasoning Curriculum for Knowledge-Preserving Parametric Tool Retrieval in Enterprise LLMs
• Source-Aware Reranking for Retrieval-Augmented Generation: A Reliability Prior Approach
• GrocLM: Grocery Category Recommendation in E-Commerce with Large Language Models
• Opti-Q: A Constraint-Based Optimization Framework for Multi-LLM Question Planning
• JKO-RAG: Distributional Retrieval as Wasserstein Free-Energy Gradient Flow
• Structure Over Scale: Schema-Constrained Causal Graphs for RAG
• Sheet As Token: A Graph-Enhanced Representation for Multi-Sheet Spreadsheet Understanding
• Controllable LLM Reasoning via Sparse Autoencoder-Based Steering
• Exploring the Security Threats of Retriever Backdoors in Retrieval-Augmented Code Generation
• Laser: Governing Long-Horizon Agentic Search via Structured Protocol and Context Register
• Towards Natural Language-Based Document Image Retrieval: New Dataset and Benchmark
• MemR³: Memory Retrieval via Reflective Reasoning for LLM Agents
• Making Large Language Models Efficient Dense Retrievers
• QuCo-RAG: Quantifying Uncertainty from the Pre-training Corpus for Dynamic Retrieval-Augmented Generation
• BanglaForge: LLM Collaboration with Self-Refinement for Bangla Code Generation
• brat: Aligned Multi-View Embeddings for Brain MRI Analysis
• External Hippocampus: Topological Cognitive Maps for Guiding Large Language Model Reasoning
• LIR³AG: A Lightweight Rerank Reasoning Strategy Framework for Retrieval-Augmented Generation
• Intelligent Knowledge Mining Framework: Bridging AI Analysis and Trustworthy Preservation
• TCDE: Topic-Centric Dual Expansion of Queries and Documents with Large Language Models for Information Retrieval
• AdaSearch: Balancing Parametric Knowledge and Search in Large Language Models via Reinforcement Learning
• ArcBERT: An LLM-based Search Engine for Exploring Integrated Multi-Omics Metadata
• Revisiting Task-Oriented Dataset Search in the Era of Large Language Models: Challenges, Benchmark, and Solution
• The Semantic Illusion: Certified Limits of Embedding-Based Hallucination Detection in RAG Systems
• IaC Generation with LLMs: An Error Taxonomy and A Study on Configuration Knowledge Injection
• Context-Picker: Dynamic context selection using multi-stage reinforcement learning
• A Simple and Effective Framework for Symmetric Consistent Indexing in Large-Scale Dense Retrieval
• Understanding Structured Financial Data with LLMs: A Case Study on Fraud Detection
• CoDA: A Context-Decoupled Hierarchical Agent with Reinforcement Learning
• How Prompts Move Language Model Behavior: Frames, Salience, and Construal as Semantic Control
• CogDoc: Towards Unified thinking in Documents
• VERAFI: Verified Agentic Financial Intelligence through Neurosymbolic Policy Generation
• Semantic Distance Measurement based on Multi-Kernel Gaussian Processes
• Citation-Grounded Code Comprehension: Preventing LLM Hallucination Through Hybrid Retrieval and Graph-Augmented Context
• Bounding Hallucinations: Information-Theoretic Guarantees for RAG Systems via Merlin-Arthur Protocols
• Replace, Don't Expand: Mitigating Context Dilution in Multi-Hop RAG via Fixed-Budget Evidence Assembly
• Cooperative Retrieval-Augmented Generation for Question Answering: Mutual Information Exchange and Ranking by Contrasting Layers
• AgriRegion: Region-Aware Retrieval for High-Fidelity Agricultural Advice
• KBQA-R1: Reinforcing Large Language Models for Knowledge Base Question Answering
• RouteRAG: Efficient Retrieval-Augmented Generation from Text and Graph via Reinforcement Learning
• Detecting Hallucinations in Graph Retrieval-Augmented Generation via Attention Patterns and Semantic Alignment
• An Index-based Approach for Efficient and Effective Web Content Extraction
• Distribution-Aware Exploration for Adaptive HNSW Search
• TopiCLEAR: Topic extraction by CLustering Embeddings with Adaptive dimensional Reduction
• Empathy by Design: Aligning Large Language Models for Healthcare Dialogue
• Optimizing Medical Question-Answering Systems: A Comparative Study of Fine-Tuned and Zero-Shot Large Language Models with RAG Framework
• The Road of Adaptive AI for Precision in Cybersecurity
• ArtistMus: A Globally Diverse, Artist-Centric Benchmark for Retrieval-Augmented Music Question Answering
• A Systematic Framework for Enterprise Knowledge Retrieval: Leveraging LLM-Generated Metadata to Enhance RAG Systems
• AdmTree: Compressing Lengthy Context with Adaptive Semantic Trees
• Automating Complex Document Workflows via Stepwise and Rollback-Enabled Operation Orchestration
• CARL: Criticality-Aware Agentic Reinforcement Learning
• On GRPO Collapse in Search-R1: The Lazy Likelihood-Displacement Death Spiral
• M3DR: Towards Universal Multilingual Multimodal Document Retrieval
• Towards Unification of Hallucination Detection and Fact Verification for Large Language Models
• Agentic Policy Optimization via Instruction-Policy Co-Evolution
• EmoRAG: Evaluating RAG Robustness to Symbolic Perturbations
• One Swallow Does Not Make a Summer: Understanding Semantic Structures in Embedding Spaces
• Bias Injection Attacks on RAG Databases and Sanitization Defenses
• Learning What Helps: Task-Aligned Context Selection for Vision Tasks
• Towards Improving Interpretability of Language Model Generation through a Structured Knowledge Discovery Approach
• Bridging the Modality Gap by Similarity Standardization with Pseudo-Positive Samples
• MegaRAG: Multimodal Knowledge Graph-Based Retrieval Augmented Generation
• MixLM: High-Throughput and Effective LLM Ranking via Text-Embedding Mix-Interaction
• R²R: A Route-to-Rerank Post-Training Framework for Multi-Domain Decoder-Only Rerankers
• Agent Discovery in Internet of Agents: Challenges and Solutions
• Stabilizing Off-Policy Training for Long-Horizon LLM Agent via Turn-Level Importance Sampling and Clipping-Triggered Normalization
• What Drives Cross-lingual Ranking? Retrieval Approaches with Multilingual Language Models
• Concept than Document: Context Compression via AMR-based Conceptual Entropy
• HyperbolicRAG: Enhancing Retrieval-Augmented Generation with Hyperbolic Representations
• Path-Constrained Retrieval: A Structural Approach to Reliable LLM Agent Reasoning Through Graph-Scoped Semantic Search
• Reuse, Don't Recompute: Efficient Large Reasoning Model Inference via Memory Orchestration
• Rethinking Retrieval: From Traditional Retrieval Augmented Generation to Agentic and Non-Vector Reasoning Systems in the Financial Domain for Large Language Models
• ENGRAM: Effective, Lightweight Memory Orchestration for Conversational Agents
• EduMod-LLM: A Modular Approach for Designing Flexible and Transparent Educational Assistants
• Learning to Compress: Unlocking the Potential of Large Language Models for Text Representation
• Comparison of Text-Based and Image-Based Retrieval in Multimodal Retrieval Augmented Generation Large Language Model Systems
• ARK: Answer-Centric Retriever Tuning via KG-augmented Curriculum Learning
• Incorporating Token Importance in Multi-Vector Retrieval
• TurkColBERT: A Benchmark of Dense and Late-Interaction Models for Turkish Information Retrieval
• MuISQA: Multi-Intent Retrieval-Augmented Generation for Scientific Question Answering
• AVATAAR: Agentic Video Answering via Temporal Adaptive Alignment and Reasoning
• CroPS: Improving Dense Retrieval with Cross-Perspective Positive Samples in Short-Video Search
• HV-Attack: Hierarchical Visual Attack for Multimodal Retrieval Augmented Generation
• ItemRAG: Item-Based Retrieval-Augmented Generation for LLM-Based Recommendation
• Noise-Robust Abstractive Compression in Retrieval-Augmented Language Models
• Attention Grounded Enhancement for Visual Document Retrieval
• Knowledge-Grounded Agentic Large Language Models for Multi-Hazard Understanding from Reconnaissance Reports
• Grounded by Experience: Generative Healthcare Prediction Augmented with Hierarchical Agentic Retrieval
• Explore More, Learn Better: Parallel MLLM Embeddings under Mutual Information Minimization
• CAT-ID²: Category-Tree Integrated Document Identifier Learning for Generative Retrieval In E-commerce
• RAGSmith: A Framework for Finding the Optimal Composition of Retrieval-Augmented Generation Methods Across Datasets
• Do LLMs and Humans Find the Same Questions Difficult? A Case Study on Japanese Quiz Answering
• CriticSearch: Fine-Grained Credit Assignment for Search Agents via a Retrospective Critic
• Towards Hyper-Efficient RAG Systems in VecDBs: Distributed Parallel Multi-Resolution Vector Search
• Thinking Forward and Backward: Multi-Objective Reinforcement Learning for Retrieval-Augmented Reasoning
• TurkEmbed: Turkish Embedding Model on NLI & STS Tasks
• DiffuGR: Generative Document Retrieval with Diffusion Language Models
• From Experience to Strategy: Empowering LLM Agents with Trainable Graph Memory
• TurkEmbed4Retrieval: Turkish Embedding Model for Retrieval Task
• Beyond Fact Retrieval: Episodic Memory for RAG with Generative Semantic Workspaces
• When, What, and How: Rethinking Retrieval-Enhanced Speculative Decoding
• Rethinking Retrieval-Augmented Generation for Medicine: A Large-Scale, Systematic Expert Evaluation and Practical Insights
• Think Before You Retrieve: Learning Test-Time Adaptive Search with Small Language Models
• Private-RAG: Answering Multiple Queries with LLMs while Keeping Your Data Private
• A Decentralized Retrieval Augmented Generation System with Source Reliabilities Secured on Blockchain
• CG-TTRL: Context-Guided Test-Time Reinforcement Learning for On-Device Large Language Models
• FLEX: Continuous Agent Evolution via Forward Learning from Experience
• A Representation Sharpening Framework for Zero Shot Dense Retrieval
• Code Review Automation using Retrieval Augmented Generation
• QueStER: Query Specification for Generative keyword-based Retrieval
• Building Specialized Software-Assistant ChatBot with Graph-Based Retrieval-Augmented Generation
• Let Me Show You: Learning by Retrieving from Egocentric Video for Robotic Manipulation
• Query Generation Pipeline with Enhanced Answerability Assessment for Financial Information Retrieval
• Search Is Not Retrieval: Decoupling Semantic Matching from Contextual Assembly in RAG
• BudgetMem: Learning Selective Memory Policies for Cost-Efficient Long-Context Processing in Language Models
• SDS KoPub VDR: A Benchmark Dataset for Visual Document Retrieval in Korean Public Documents
• E-CARE: An Efficient LLM-based Commonsense-Augmented Framework for E-Commerce
• Abductive Inference in Retrieval-Augmented Language Models: Generating and Validating Missing Premises
• GEMMA-SQL: A Novel Text-to-SQL Model Based on Large Language Models
• Cache Mechanism for Agent RAG Systems
• MemSearcher: Training LLMs to Reason, Search and Manage Memory via End-to-End Reinforcement Learning
• Beyond Single Embeddings: Capturing Diverse Targets with Multi-Query Retrieval
• Trove: A Flexible Toolkit for Dense Retrieval
• PROPEX-RAG: Enhanced GraphRAG using Prompt-Driven Prompt Execution
• Towards LLM-Powered Task-Aware Retrieval of Scientific Workflows for Galaxy
• Rescuing the Unpoisoned: Efficient Defense against Knowledge Corruption Attacks on RAG Systems
• OceanAI: A Conversational Platform for Accurate, Transparent, Near-Real-Time Oceanographic Insights
• Separate the Wheat from the Chaff: Winnowing Down Divergent Views in Retrieval Augmented Generation
• Taxonomy-based Negative Sampling In Personalized Semantic Search for E-commerce
• LIR: The First Workshop on Late Interaction and Multi Vector Retrieval @ ECIR 2026
• A Survey on Deep Text Hashing: Efficient Semantic Text Retrieval with Binary Representation
• Adapting Large Language Models to Emerging Cybersecurity using Retrieval Augmented Generation
• AgentBnB: A Browser-Based Cybersecurity Tabletop Exercise with Large Language Model Support and Retrieval-Aligned Scaffolding
• MARAG-R1: Beyond Single Retriever via Reinforcement-Learned Multi-Tool Agentic Retrieval
• InfoFlow: Reinforcing Search Agent Via Reward Density Optimization
• Polybasic Speculative Decoding Through a Theoretical Perspective
• Towards Global Retrieval Augmented Generation: A Benchmark for Corpus-Level Reasoning
• FARSIQA: Faithful and Advanced RAG System for Islamic Question Answering
• Generalized Pseudo-Relevance Feedback
• GAP: Graph-Based Agent Planning with Parallel Tool Use and Reinforcement Learning
• BM25 Wins at Scale: A Scaling Study of Retrieval-Augmented Generation Paradigms
• SimpleWikiSearch: A Clean Offline Wikipedia Environment for Agentic Search
• Tools Are Not Islands: Set-Level Tool Retrieval for LLM Agents via Query-Conditioned Hyperedge Prediction
• Finding Diamonds in Conversation Haystacks: A Benchmark for Conversational Data Retrieval
• Assessing early oil industry awareness of the impacts of fossil fuels on coral reefs using a novel AI agent
• Is Dimensionality a Barrier for Retrieval Models?
• Understanding Wacky Weights: A Dissection of SPLADE's Learned Term Importance
• POINTS-Seeker: An Open Recipe for Multimodal Search Agents with Visual Memory Management
• MathNet: a Global Multimodal Benchmark for Mathematical Reasoning and Retrieval
• Exploring the Intersection of AI, Language, and Law: A Bibliometric Analysis
• MICA: Multi-granularity Intertemporal Credit Assignment for Long-Horizon Emotional Support Dialogue
• Can Knowledge-Graph-based Retrieval Augmented Generation Really Retrieve What You Need?
• Sharpness-Guided Group Relative Policy Optimization via Probability Shaping
• Secure Retrieval-Augmented Generation against Poisoning Attacks
• Iterative Critique-Refine Framework for Enhancing LLM Personalization
• Repurposing Synthetic Data for Fine-grained Search Agent Supervision
• Eigenfunction Extraction for Ordered Representation Learning
• Optimizing Retrieval for RAG via Reinforced Contrastive Learning
• Talk2Ref: A Dataset for Reference Prediction from Scientific Talks
• Mitigating Hallucination in Large Language Models (LLMs): An Application-Oriented Survey on RAG, Reasoning, and Agentic Systems
• Metadata-Driven Retrieval-Augmented Generation for Financial Question Answering
• Minimizing Human Intervention in Online Classification
• Agentic Meta-Orchestrator for Multi-task Copilots
• Multi-Modal Fact-Verification Framework for Reducing Hallucinations in Large Language Models
• Contextual Tokenization for Graph Inverted Indices
• E2Rank: Your Text Embedding can Also be an Effective and Efficient Listwise Reranker
• FAIR-RAG: Faithful Adaptive Iterative Refinement for Retrieval-Augmented Generation
• LSPRAG: LSP-Guided RAG for Language-Agnostic Real-Time Unit Test Generation
• Multimodal Item Scoring for Natural Language Recommendation via Gaussian Process Regression with LLM Relevance Judgments
• Large Language Models Meet Text-Attributed Graphs: A Survey of Integration Frameworks and Applications
• GlobalRAG: Enhancing Global Reasoning in Multi-hop Question Answering via Reinforcement Learning
• GRATING: Low-Latency and Memory-Efficient Semantic Selection on Device
• Multimedia-Aware Question Answering: A Review of Retrieval and Cross-Modal Reasoning Architectures
• Think Parallax: Solving Multi-Hop Problems via Multi-View Knowledge-Graph-Based Retrieval-Augmented Generation
• From Masks to Worlds: A Hitchhiker's Guide to World Models
• ToolDreamer: Instilling LLM Reasoning Into Tool Retrievers
• CoRECT: A Framework for Evaluating Embedding Compression Techniques at Scale
• C2T-ID: Converting Semantic Codebooks to Textual Document Identifiers for Generative Search
• Search Self-play: Pushing the Frontier of Agent Capability without Supervision
• Sherlock Your Queries: Learning to Ask the Right Questions for Dialogue-Based Retrieval
• LLMs as Sparse Retrievers:A Framework for First-Stage Product Search
• MENTOR: A Reinforcement Learning Framework for Enabling Tool Use in Small Models via Teacher-Optimized Rewards
• Position: LLM Watermarking Should Align Stakeholders' Incentives for Practical Adoption
• AcademicEval: Live Long-Context LLM Benchmark
• Agentic Reinforcement Learning for Search Misaligns Instruction-Tuning
• Towards Mixed-Modal Retrieval for Universal Retrieval-Augmented Generation
• DSEBench: A Test Collection for Explainable Dataset Search with Examples
• OG-Rank: Learning to Rank Fast and Slow with Uncertainty and Reward-Trend Guided Adaptive Exploration
• Towards Context-aware Reasoning-enhanced Generative Searching in E-commerce
• Exact Nearest-Neighbor Search on Energy-Efficient FPGA Devices
• Cross-Genre Authorship Attribution via LLM-Based Retrieve-and-Rerank
• Mixture of Experts Approaches in Dense Retrieval Tasks
• DMRetriever: A Family of Models for Improved Text Retrieval in Disaster Management
• Hierarchical Semantic Retrieval with Cobweb
• Rewriting History: A Recipe for Interventional Analyses to Study Data Effects on Model Behavior
• Towards Agentic Self-Learning LLMs in Search Environment
• Large Scale Retrieval for the LinkedIn Feed using Causal Language Models
• JEDA: Query-Free Clinical Order Search from Ambient Dialogues
• Knowledge-based Visual Question Answer with Multimodal Processing, Retrieval and Filtering
• BRIEF-Pro: Universal Context Compression with Short-to-Long Synthesis for Fast and Accurate Multi-Hop Reasoning
• Big Reasoning with Small Models: Instruction Retrieval at Inference Time
• ReMindRAG: Low-Cost LLM-Guided Knowledge Graph Traversal for Efficient RAG
• Grounding Long-Context Reasoning with Contextual Normalization for Retrieval-Augmented Generation
• Retrieval-in-the-Chain: Bootstrapping Large Language Models for Generative Retrieval
• LLM-guided Hierarchical Search for End-to-end Reasoning Intensive Retrieval
• PIShield: Detecting Prompt Injection Attacks via Intrinsic LLM Features
• CTRL-Rec: Controlling Recommender Systems With Natural Language
• Understanding Parametric Knowledge Injection in Retrieval-Augmented Generation
• Probing Latent Knowledge Conflict for Faithful Retrieval-Augmented Generation
• SAIL-Embedding Technical Report: Omni-modal Embedding Foundation Model
• VizCopilot: Fostering Appropriate Reliance on Enterprise Chatbots with Context Visualization
• REGENT: Relevance-Guided Attention for Entity-Aware Multi-Vector Neural Re-Ranking
• QDER: Query-Specific Document and Entity Representations for Multi-Vector Document Re-Ranking
• Uncertainty Quantification for Retrieval-Augmented Reasoning
• LLM-Specific Utility: A New Perspective for Retrieval-Augmented Generation
• Domain-Specific Data Generation Framework for RAG Adaptation
• AccurateRAG: A Framework for Building Accurate Retrieval-Augmented Question-Answering Applications
• Scalable and Explainable Enterprise Knowledge Discovery Using Graph-Centric Hybrid Retrieval
• VeritasFi: An Adaptable, Multi-tiered RAG Framework for Multi-modal Financial Question Answering
• Agentic RAG for Software Testing with Hybrid Vector-Graph and Multi-Agent Orchestration
• DRIFT: Decompose, Retrieve, Illustrate, then Formalize Theorems
• BrowserAgent: Building Web Agents with Human-Inspired Web Browsing Actions
• Taming a Retrieval Framework to Read Images in Humanlike Manner for Augmenting Generation of MLLMs
• ZeroGR: A Generalizable and Scalable Framework for Zero-Shot Generative Retrieval
• Beyond the limitation of a single query: Train your LLM for query expansion with Reinforcement Learning
• Learning from AVA: Early Lessons from a Curated and Trustworthy Generative AI for Policy and Development Research
• Text2Token: Unsupervised Text Representation Learning with Token Target Prediction
• DSPO: Stable and Efficient Policy Optimization for Agentic Search and Reasoning
• Autoencoding-Free Context Compression for LLMs via Contextual Semantic Anchors
• EcphoryRAG: Re-Imagining Knowledge-Graph RAG via Human Associative Memory
• GRETEL: A Goal-driven Retrieval and Execution-based Trial Framework for LLM Tool Selection Enhancing
• NL2GenSym: Natural Language to Generative Symbolic Rules for SOAR Cognitive Architecture via Large Language Models
• PairSem: LLM-Guided Pairwise Semantic Matching for Scientific Document Retrieval
• TaoSR-AGRL: Adaptive Guided Reinforcement Learning Framework for E-commerce Search Relevance
• HiPRAG: Hierarchical Process Rewards for Efficient Agentic Retrieval Augmented Generation
• Multilingual Generative Retrieval via Cross-lingual Semantic Compression
• MemWeaver: A Hierarchical Memory from Textual Interactive Behaviors for Personalized Generation
• Struc-EMB: The Potential of Structure-Aware Encoding in Language Embeddings
• QAgent: A modular Search Agent with Interactive Query Understanding
• VersionRAG: Version-Aware Retrieval-Augmented Generation for Evolving Documents
• Lean Finder: Semantic Search for Mathlib That Understands User Intents
• Haystack Engineering: Context Engineering for Heterogeneous and Agentic Long-Context Evaluation
• LAD-RAG: Layout-aware Dynamic RAG for Visually-Rich Document Understanding
• Table Question Answering in the Era of Large Language Models: A Comprehensive Survey of Tasks, Methods, and Evaluation
• Search-R3: Unifying Reasoning and Embedding in Large Language Models
• SoftMatcha 2: A Fast and Soft Pattern Matcher for Trillion-Scale Corpora
• Study on LLMs for Promptagator-Style Dense Retriever Training
• Reproducing and Extending Causal Insights Into Term Frequency Computation in Neural Rankers
• ToolMem: Enhancing Multimodal Agents with Learnable Tool Capability Memory
• Adaptive Tool Generation with Models as Tools and Reinforcement Learning
• Valid Stopping for LLM Generation via Empirical Dynamic Formal Lift
• Stratified GRPO: Handling Structural Heterogeneity in Reinforcement Learning of LLM Search Agents
• DecEx-RAG: Boosting Agentic Retrieval-Augmented Generation with Decision and Execution Optimization via Process Supervision
• Scalable In-context Ranking with Generative Models
• Guided Query Refinement: Multimodal Hybrid Retrieval with Test-Time Optimization
• ModernBERT + ColBERT: Enhancing biomedical RAG through an advanced re-ranking retriever
• Contrastive Retrieval Heads Improve Attention-Based Re-Ranking
• Fine-grained auxiliary learning for real-world product recommendation
• GRACE: Generative Representation Learning via Contrastive Policy Optimization
• Improving Consistency in Retrieval-Augmented Systems with Group Similarity Rewards
• MetaFind: Scene-Aware 3D Asset Retrieval for Coherent Metaverse Scene Generation
• SECA: Semantically Equivalent and Coherent Attacks for Eliciting LLM Hallucinations
• JEF-Hinter: Leveraging Offline Knowledge for Improving Web Agents Adaptation
• Equipping Retrieval-Augmented Large Language Models with Document Structure Awareness
• EvoEngineer: Mastering Automated CUDA Kernel Code Evolution with Large Language Models
• Less LLM, More Documents: Searching for Improved RAG
• Geolog-IA: Conversational System for Academic Theses
• StepChain GraphRAG: Reasoning Over Knowledge Graphs for Multi-Hop Question Answering
• A Simple but Effective Elaborative Query Reformulation Approach for Natural Language Recommendation
• ModernVBERT: Towards Smaller Visual Document Retrievers
• MASH: Modeling Abstention via Selective Help-Seeking
• ReSeek: A Self-Correcting Framework for Search Agents with Instructive Rewards
• TokMem: Tokenized Procedural Memory for Large Language Models
• Erase to Improve: Erasable Reinforcement Learning for Search-Augmented LLMs
• From Factoid Questions to Data Product Requests: Benchmarking Data Product Discovery over Tables and Text
• Learning to Route: A Rule-Driven Agent Framework for Hybrid-Source Retrieval-Augmented Generation
• Optimizing What Matters: AUC-Driven Learning for Robust Neural Retrieval
• Fairness Testing in Retrieval-Augmented Generation: How Small Perturbations Reveal Bias in Small Language Models
• MR²-Bench: Going Beyond Matching to Reasoning in Multimodal Retrieval
• TruthRL: Incentivizing Truthful LLMs via Reinforcement Learning
• SafeMind: Benchmarking and Mitigating Safety Risks in Embodied LLM Agents
• Hybrid Reward Normalization for Process-supervised Non-verifiable Agentic Tasks
• SimulRAG: Simulator-based RAG for Grounding LLMs in Long-form Scientific QA
• Investigating Language and Retrieval Bias in Multilingual Previously Fact-Checked Claim Detection
• Retro*: Optimizing LLMs for Reasoning-Intensive Document Retrieval
• UniDex: Rethinking Search Inverted Indexing with Unified Semantic Modeling
• Let LLMs Speak Embedding Languages: Generative Text Embeddings via Iterative Contrastive Refinement
• DocPruner: A Storage-Efficient Framework for Multi-Vector Visual Document Retrieval via Adaptive Patch-Level Embedding Pruning
• Investigating Multi-layer Representations for Dense Passage Retrieval
• Transformer Tafsir at QIAS 2025 Shared Task: Hybrid Retrieval-Augmented Generation for Islamic Knowledge Question Answering
• DiffuSpec: Unlocking Diffusion Language Models for Speculative Decoding
• RANGER -- Repository-Level Agent for Graph-Enhanced Retrieval
• Your RAG is Unfair: Exposing Fairness Vulnerabilities in Retrieval-Augmented Generation via Backdoor Attacks
• Does Generative Retrieval Overcome the Limitations of Dense Retrieval?
• Effect of Model Merging in Domain-Specific Ad-hoc Retrieval
• Beyond RAG vs. Long-Context: Learning Distraction-Aware Retrieval for Efficient Knowledge Grounding
• QuantMind: A Context-Engineering Based Knowledge Framework for Quantitative Finance
• Tree Search for LLM Agent Reinforcement Learning
• CLAUSE: Agentic Neuro-Symbolic Knowledge Graph Reasoning via Dynamic Learnable Context Engineering
• An LLM-based Agentic Framework for Accessible Network Control
• Learning Contextual Retrieval for Robust Conversational Search
• GLM-RAG: Graph Language Models for Graph-Based Retrieval-Augmented Generation
• MemTxn: A Transaction Boundary for Source-Supported Updates and Complete-State Recovery in Agent Memory
• DeepResearch Agent System
• EMBL AI Librarian: Life-Sciences Knowledge Layer for AI Agents
• ConMem: Contribution-Aware Memory for Long-Horizon Manufacturing Inspection Logs
• LayerRAG-Bench: A Cross-Layer Reliability Benchmark for Agentic Retrieval-Augmented Generation
• Dr-DCI: Scaling Direct Corpus Interaction via Dynamic Workspace Expansion
• Retrieval from Within: An Intrinsic Capability of Attention-Based Models
• LLM2Vec-Gen: Generative Embeddings from Large Language Models
• KuaiSearch: An E-Commerce Search Dataset with Authentic Queries and Product Texts for Recall, Ranking, and Relevance
• LexSemBridge: Fine-Grained Dense Representation Enhancement through Token-Aware Embedding Augmentation
• Randomly Removing 50% of Dimensions in Text Embeddings has Minimal Impact on Retrieval and Classification Tasks
• Memory in Large Language Models: Mechanisms, Evaluation and Evolution
• Semantic Search for Information Retrieval
• Financial Risk Relation Identification through Dual-view Adaptation
• Confidence-Aware Routing for Large Language Model Reliability Enhancement: A Multi-Signal Approach to Pre-Generation Hallucination Mitigation
• AIRwaves at CheckThat! 2025: Retrieving Scientific Sources for Implicit Claims on Social Media with Dual Encoders and Neural Re-Ranking
• A Generative Framework for Personalized Sticker Retrieval
• AttnComp: Attention-Guided Adaptive Context Compression for Retrieval-Augmented Generation
• Improving Zero-shot Sentence Decontextualisation with Content Selection and Planning
• Individualized non-uniform quantization for vector search
• GRIL: Knowledge Graph Retrieval-Integrated Learning with Large Language Models
• The Role of Vocabularies in Learning Sparse Representations for Ranking
• Evaluating the Effectiveness and Scalability of LLM-Based Data Augmentation for Retrieval
• Enhancing Financial RAG with Agentic AI and Multi-HyDE: A Novel Approach to Knowledge Retrieval and Hallucination Reduction
• Efficient and Versatile Model for Multilingual Information Retrieval of Islamic Text: Development and Deployment in Real-World Scenarios
• Quantifying Self-Awareness of Knowledge in Large Language Models
• AIP: Subverting Retrieval-Augmented Generation via Adversarial Instructional Prompt
• What's the Best Way to Retrieve Slides? A Comparative Study of Multimodal, Caption-Based, and Hybrid Retrieval Techniques
• Retrieval Capabilities of Large Language Models Scale with Pretraining FLOPs
• Linguistic Nepotism: Trading-off Quality for Language Preference in Multilingual RAG
• HSGM: Hierarchical Segment-Graph Memory for Scalable Long-Text Semantics
• Who Taught the Lie? Responsibility Attribution for Poisoned Knowledge in Retrieval-Augmented Generation
• Modernizing Facebook Scoped Search: Keyword and Embedding Hybrid Retrieval with LLM Evaluation
• Omne-R1: Learning to Reason with Memory for Multi-hop Question Answering
• CORE-RAG: Lossless Compression for Retrieval-Augmented LLMs via Reinforcement Learning
• Conan-Embedding-v2: Training an LLM from Scratch for Text Embeddings
• ScaleDoc: Scaling LLM-based Predicates over Large Document Collections
• Exposing Privacy Risks in Graph Retrieval-Augmented Generation
• InfoGain-RAG: Boosting Retrieval-Augmented Generation via Document Information Gain-based Reranking and Filtering
• MA-DPR: Manifold-aware Distance Metrics for Dense Passage Retrieval
• Zero-shot Multimodal Document Retrieval via Cross-modal Question Generation
• Graph-Enhanced Retrieval-Augmented Question Answering for E-Commerce Customer Support
• Self-Evolving LLMs via Continual Instruction Tuning
• AI Answer Engine Citation Behavior An Empirical Analysis of the GEO16 Framework
• A Survey on Retrieval And Structuring Augmented Generation with Large Language Models
• DeAR: Dual-Stage Document Reranking with Reasoning Agents via LLM Distillation
• ReFactX: Scalable Reasoning with Reliable Facts via Constrained Generation
• Bridging the Capability Gap: Joint Alignment Tuning for Harmonizing LLM-based Multi-Agent Systems
• SEDM: Scalable Self-Evolving Distributed Memory for Agents
• Boosting Data Utilization for Multilingual Dense Retrieval
• IMDMR: An Intelligent Multi-Dimensional Memory Retrieval System for Enhanced Conversational AI
• LLM Ensemble for RAG: Role of Context Length in Zero-Shot Question Answering for BioASQ Challenge
• Adversarial Attacks Against Automated Fact-Checking: A Survey
• Handling Open-Vocabulary Constructs in Formalizing Specifications: Retrieval-Augmented Parsing with Expert Knowledge
• ImportSnare: Directed "Code Manual" Hijacking in Retrieval-Augmented Code Generation
• A Survey of Long-Document Retrieval in the PLM and LLM Era
• Multi-view-guided Passage Reranking with Large Language Models
• PatchSeeker: Mapping NVD Records to their Vulnerability-fixing Commits with LLM Generated Commits and Embeddings
• Beyond Sequential Reranking: Reranker-Guided Search Improves Reasoning Intensive Retrieval
• Reason to Retrieve: Enhancing Query Understanding through Decomposition and Interpretation
• Domain-Aware RAG: MoL-Enhanced RL for Efficient Training and Scalable Retrieval
• Language Bias in Information Retrieval: The Nature of the Beast and Mitigation Methods
• DecoupleSearch: Decouple Planning and Search via Hierarchical Reward Modeling
• DCMI: A Differential Calibration Membership Inference Attack Against Retrieval-Augmented Generation
• Understanding the Influence of Synthetic Data for Text Embedders
• Toward Efficient and Scalable Design of In-Memory Graph-Based Vector Search
• KERAG: Knowledge-Enhanced Retrieval-Augmented Generation for Advanced Question Answering
• NER Retriever: Zero-Shot Named Entity Retrieval with Type-Aware Embeddings
• LLM-based Relevance Assessment for Web-Scale Search Evaluation at Pinterest
• OPERA: A Reinforcement Learning--Enhanced Orchestrated Planner-Executor Architecture for Reasoning-Oriented Multi-Hop Retrieval
• Retrieval-Augmented Defense: Adaptive and Controllable Jailbreak Prevention for Large Language Models
• Training LLMs to be Better Text Embedders through Bidirectional Reconstruction
• Lighting the Way for BRIGHT: Reproducible Baselines with Anserini, Pyserini, and RankLLM
• CMRAG: Co-modality-based visual document retrieval and question answering
• First RAG, Second SEG: A Training-Free Paradigm for Camouflaged Object Detection
• Upcycling Candidate Tokens of Large Language Models for Query Expansion
• Hierarchical Vision-Language Reasoning for Multimodal Multiple-Choice Question Answering
• Inducing Faithfulness in Structured Reasoning via Counterfactual Sensitivity
• XLQA: A Benchmark for Locale-Aware Multilingual Open-Domain Question Answering
• Beyond the Surface: A Solution-Aware Retrieval Model for Competition-level Code Generation
• Privacy-Preserving Reasoning with Knowledge-Distilled Parametric Retrieval Augmented Generation
• VerlTool: Towards Holistic Agentic Reinforcement Learning with Tool Use
• Re3: Learning to Balance Relevance & Recency for Temporal Information Retrieval
• Negative Matters: Multi-Granularity Hard-Negative Synthesis and Anchor-Token-Aware Pooling for Enhanced Text Embeddings
• Multimodal Iterative RAG for Knowledge-Intensive Visual Question Answering
• Energy Landscapes Enable Reliable Abstention in Retrieval-Augmented Large Language Models for Healthcare
• EviNote-RAG: Enhancing RAG Models via Answer-Supportive Evidence Notes
• L3Cube-MahaSTS: A Marathi Sentence Similarity Dataset and Models
• KG-CQR: Leveraging Structured Relation Representations in Knowledge Graphs for Contextual Query Retrieval
• AI-SearchPlanner: Modular Agentic Search via Pareto-Optimal Multi-Objective Reinforcement Learning
• SEAL: Structure and Element Aware Learning to Improve Long Structured Document Retrieval
• SurGE: A Benchmark and Evaluation Framework for Scientific Survey Generation
• Fact or Facsimile? Evaluating the Factual Robustness of Modern Retrievers
• Can Compact Language Models Search Like Agents? Distillation-Guided Policy Optimization for Preserving Agentic RAG Capabilities
• From Search to Reasoning: A Five-Level RAG Capability Framework for Enterprise Data
• Exploring Selective Retrieval-Augmentation for Long-Tail Legal Text Classification
• NLKI: A lightweight Natural Language Knowledge Integration Framework for Improving Small VLMs in Commonsense VQA Tasks
• Uncovering the Bigger Picture: Comprehensive Event Understanding Via Diverse News Retrieval
• MODE: Mixture of Document Experts for RAG
• AI for Statutory Simplification: A Comprehensive State Legal Corpus and Labor Benchmark
• ArgRAG: Explainable Retrieval Augmented Generation using Quantitative Bipolar Argumentation
• Test-time Corpus Feedback: From Retrieval to RAG
• Optimization of Latent-Space Compression using Game-Theoretic Techniques for Transformer-Based Vector Search
• UniC-RAG: Universal Knowledge Corruption Attacks to Retrieval-Augmented Generation
• Improving End-to-End Training of Retrieval-Augmented Generation Models via Joint Stochastic Approximation
• Test-Time Scaling Strategies for Generative Retrieval in Multimodal Conversational Recommendations
• HyST: LLM-Powered Hybrid Retrieval over Semi-Structured Tabular Data
• Identifying and Answering Questions with False Assumptions: An Interpretable Approach
• FLAIR: Feedback Learning for Adaptive Information Retrieval
• CardAIc-Agents: A Multimodal Framework with Hierarchical Adaptation for Cardiac Care Support
• Ontology-Guided Query Expansion for Biomedical Document Retrieval using Large Language Models
• CoDiEmb: A Collaborative yet Distinct Framework for Unified Representation Learning in Information Retrieval and Semantic Textual Similarity
• Retrieval-augmented reasoning with lean language models
• RAG for Geoscience: What We Expect, Gaps and Opportunities
• ALAS: Autonomous Learning Agent for Self-Updating Language Models
• LeanRAG: Knowledge-Graph-Based Generation with Semantic Aggregation and Hierarchical Retrieval
• RAGulating Compliance: A Multi-Agent Knowledge Graph for Regulatory QA
• SYNAPSE-G: Bridging Large Language Models and Graph Learning for Rare Event Classification
• Improving Dense Passage Retrieval with Multiple Positive Passages
• ParallelSearch: Train your LLMs to Decompose Query and Search Sub-queries in Parallel with Reinforcement Learning
• SMA: Who Said That? Auditing Membership Leakage in Semi-Black-box RAG Controlling
• Improving Document Retrieval Coherence for Semantically Equivalent Queries
• Beyond Ten Turns: Unlocking Long-Horizon Agentic Search with Large-Scale Asynchronous RL
• AURA: A Fine-Grained Benchmark and Decomposed Metric for Audio-Visual Reasoning
• CLAP: Coreference-Linked Augmentation for Passage Retrieval
• BiXSE: Improving Dense Retrieval via Probabilistic Graded Relevance Distillation
• Synthesizing scientific literature with retrieval-augmented language models
• Beyond Perplexity: Let the Reader Select Retrieval Summaries via Spectrum Projection Score
• Role of Large Language Models and Retrieval-Augmented Generation for Accelerating Crystalline Material Discovery: A Systematic Review
• Improving Table Retrieval with Question Generation from Partial Tables
• BrowseComp-Plus: A More Fair and Transparent Evaluation Benchmark of Deep-Research Agent
• CoCoLex: Confidence-guided Copy-based Decoding for Grounded Legal Text Generation
• RankArena: A Unified Platform for Evaluating Retrieval, Reranking and RAG with Human and LLM Feedback
• LAG: Logic-Augmented Generation from a Cartesian Perspective
• RRRA: Resampling and Reranking through a Retriever Adapter
• BEE-RAG: Balanced Entropy Engineering for Retrieval-Augmented Generation
• FinAgentBench: A Benchmark Dataset for Agentic Retrieval in Financial Question Answering
• PAIRS: Parametric-Verified Adaptive Information Retrieval and Selection for Efficient RAG
• A Few Words Can Distort Graphs: Knowledge Poisoning Attacks on Graph-based Retrieval-Augmented Generation of Large Language Models
• An Entity Linking Agent for Question Answering
• AttnTrace: Attention-based Context Traceback for Long-Context LLMs
• LLMDistill4Ads: Using Cross-Encoders to Distill from LLM Signals for Advertiser Keyphrase Recommendations
• PyLate: Flexible Training and Retrieval for Late Interaction Models
• Defending Against Knowledge Poisoning Attacks During Retrieval-Augmented Generation
• Beyond Chunks and Graphs: Retrieval-Augmented Generation through Triplet-Driven Thinking
• ReMoMask: Retrieval-Augmented Masked Motion Generation
• CoCoA: Collaborative Chain-of-Agents for Parametric-Retrieved Knowledge Synergy
• Balancing the Blend: An Experimental Analysis of Trade-offs in Hybrid Search
• From Generator to Embedder: Harnessing Innate Abilities of Multimodal LLMs via Building Zero-Shot Discriminative Embedding Model
• MAO-ARAG: Multi-Agent Orchestration for Adaptive Retrieval-Augmented Generation
• Automating AI Failure Tracking: Semantic Association of Reports in AI Incident Database
• From Static to Dynamic: A Streaming RAG Approach to Real-time Knowledge Base
• Enhanced Arabic Text Retrieval with Attentive Relevance Scoring
• Causal2Vec: Improving Decoder-only LLMs as Versatile Embedding Models
• MUST-RAG: MUSical Text Question Answering with Retrieval Augmented Generation
• Bidirectional Likelihood Estimation with Multi-Modal Large Language Models for Text-Video Retrieval
• Annotation-Free Reinforcement Learning Query Rewriting via Verifiable Search Reward
• FACTORY: A Challenging Human-Verified Prompt Set for Long-Form Factuality

Related

• RocketQA: An Optimized Training Approach to Dense Passage Retrieval for Open-Domain Question Answering
• Open Domain Question Answering over Tables via Dense Retrieval
• Efficient Passage Retrieval with Hashing for Open-domain Question Answering
• A Replication Study of Dense Passage Retriever
• A Memory Efficient Baseline for Open Domain Question Answering