Retrieval-Augmented Generation

Also known as RAG

Retrieval-augmented generation (RAG) studies how to give language models relevant information from documents, databases, or other sources at answer time, improving access to knowledge that may be missing or outdated in model parameters. Common approaches combine dense or sparse search with hybrid retrieval, reranking, and document chunking; GraphRAG and agentic methods add structure or iterative search for multi-hop questions and long documents. Work spans open-domain QA benchmarks as well as specialized settings such as legal, medical, code, and multimodal retrieval. A central tension is that more retrieved context can improve evidence coverage but also introduce noise, conflicts, cost, and unsupported answers, so systems increasingly focus on selective retrieval, verification, citation fidelity, and realistic evaluation.

Latest papers

796 papers72 papers in the last four weeks, down 14% on the four weeks before. 0.4% of all new papers.

IG-Search: Step-Level Information Gain Rewards for Search-Augmented Reasoning
Apr 16, 2026Zihan Liang, Yufei Ma, Ben Chen +6Retrieval-Augmented Generation RL for Language Model Reasoning
Kuaishou Technology
Enhancing Mental Health Counseling Support in Bangladesh using Culturally-Grounded Knowledge
Apr 16, 2026Md Arid Hasan, Azhagu Meena SP, Aditya Khan +6Retrieval-Augmented Generation Mental Health Counseling
University of Toronto, Canada · mPower Social Enterprises Ltd, Bangladesh · National Institute of Mental Health, Bangladesh +2
Don't Retrieve, Navigate: Distilling Enterprise Knowledge into Navigable Agent Skills for QA and RAG
Apr 16, 2026Yiqun Sun, Pengfei Wei, Lawrence B. HsiehRetrieval-Augmented Generation LLM Agent Skill Learning
Magellan Technology Research Institute (MTRI)
NaviRAG: Towards Active Knowledge Navigation for Retrieval-Augmented Generation
Apr 14, 2026Jihao Dai, Dingjun Wu, Yuxuan Chen +4Retrieval-Augmented Generation Agentic RAG
Tsinghua University · Nanjing University · Northeastern University
Transforming External Knowledge into Triplets for Enhanced Retrieval in RAG of LLMs
Apr 14, 2026Xudong Wang, Chaoning Zhang, Qigan Sun +7Structured RAG Retrieval-Augmented Generation
School of Computing, Kyung Hee University, Yongin-si, South Korea · School of Computer Science and Engineering, University of Electronic Science and Technology of China, Chengdu, China · School of Information and Software Engineering, University of Electronic Science and Technology of China, Chengdu, China +2
ClusterRAG: Cluster-Based Collaborative Filtering for Personalized Retrieval-Augmented Generation
Apr 14, 2026Gibson Nkhata, Uttamasha Anjally Oyshi, Quan Mai +1Density-Based Clustering Retrieval-Augmented Generation
University of Arkansas, Fayetteville, AR 72701, USA · Walmart Inc., Bentonville, AR 72716, USA
Knowledge Is Not Static: Order-Aware Hypergraph RAG for Language Models
Apr 14, 2026Keshu Wu, Chenchen Kuai, Zihao Li +6Hypergraph-Based RAG Retrieval-Augmented Generation
Texas A&M University · University of Illinois Urbana-Champaign · The University of Kansas
Retrieval-Augmented Generation Must Move Beyond Factual Grounding to Represent Diverse Opinions
Apr 13, 2026Aditya Agrawal, Alwarappan Nakkiran, Aman Singh Thakur +2Retrieval-Augmented Generation Aleatoric Uncertainty
Retrieval Is Not Enough: Why Organizational AI Needs Epistemic Infrastructure
Apr 13, 2026Federico Bottino, Carlo Ferrero, Nicholas Dosio +1Knowledge Representation Epistemic Uncertainty
Kakashi Ventures Accelerator (KVA) · Massachusetts Institute of Technology
Legal2LogicICL: Improving Generalization in Transforming Legal Cases to Logical Formulas via Diverse Few-Shot Learning
Apr 13, 2026Jieying Xue, Phuong Minh Nguyen, Ha Thanh Nguyen +2Legal Reasoning Legal NLP
Center for Juris-Informatics, ROIS-DS, Tokyo, Japan · Japan Advanced Institute of Science and Technology, Ishikawa, Japan
Retrieval as Generation: A Unified Framework with Self-Triggered Information Planning
Apr 13, 2026Bo Li, Mingda Wang, Gexiang Fang +2Retrieval-Augmented Generation Adaptive RAG
National Engineering Research Center for Software Engineering, Peking University · School of Computer Science, Peking University · School of Health Sciences and Biomedical Engineering, Hebei University of Technology
Exploring Knowledge Conflicts for Faithful LLM Reasoning: Benchmark and Method
Apr 13, 2026Tianzhe Zhao, Jiaoyan Chen, Shuxiu Zhang +3Knowledge Conflicts in Language Models LLM Reasoning with Graphs
School of Computer Science and Technology, Xi’an Jiaotong University · Department of Computer Science, The University of Manchester · Hunan University +1
RAG-KT: Cross-platform Explainable Knowledge Tracing with Multi-view Fusion Retrieval Generation
Apr 13, 2026Zhiyi Duan, Hongyu Yuan, Rui LiuKnowledge Tracing AI in Education
Inner Mongolia University, Hohhot, China
AOP-Smart: A RAG-Enhanced Large Language Model Framework for Adverse Outcome Pathway Analysis
Apr 13, 2026Qinjiang Niu, Lu YanLLM Hallucination Mitigation Retrieval-Augmented Generation
Nanyang Normal University, Nanyang, China
Self-Correcting RAG: Enhancing Faithfulness via MMKP Context Selection and NLI-Guided MCTS
Apr 12, 2026Shijia Xu, Zhou Wu, Xiaolong Jia +3Retrieval-Augmented Generation RAG Robustness
Chongqing University, China · Queen Mary University of London, UK · Chongqing Key Laboratory of Big Data Intelligence and Privacy Computing, China +1
Structure-Grounded Knowledge Retrieval via Code Dependencies for Multi-Step Data Reasoning
Apr 12, 2026Xinyi HuangKnowledge Augmentation for Language Models Retrieval-Augmented Generation
Simon Fraser University
CodaRAG: Connecting the Dots with Associativity Inspired by Complementary Learning
Apr 12, 2026Cheng-Yen Li, Xuanjun Chen, Claire Lin +4Retrieval-Augmented Generation Graph Retrieval-Augmented Generation
National Taiwan University, Taipei, Taiwan · NTU Artificial Intelligence Center of Research Excellence, National Taiwan University.
BLUEmed: Retrieval-Augmented Multi-Agent Debate for Clinical Error Detection
Apr 12, 2026Saukun Thika You, Nguyen Anh Khoa Tran, Wesley K. Marizane +3Language Model Error Detection Multi-Agent Debate
Department of Computer Science, University of Memphis, Memphis, United States · Department of Computer Information Systems, Alabama State University, Montgomery, United States
Neuro-Oracle: A Trajectory-Aware Agentic RAG Framework for Interpretable Epilepsy Surgical Prognosis
Apr 10, 2026Aizierjiang Aiersilan, Mohamad KoubeissiClinical Outcome Prediction Retrieval-Augmented Generation
The George Washington University
PriHA: A RAG-Enhanced LLM Framework for Primary Healthcare Assistant in Hong Kong
Apr 10, 2026Richard Wai Cheung Chan, Shanru Lin, Ya-nan Ma +3Medical LLMs Healthcare
The Hong Kong Polytechnic University, Hong Kong, China · Haikou Affiliated Hospital of Central South University Xiangya School of Medicine, Haikou, China · Sun Yat-Sen University Cancer Center, Guangzhou, China
Micro-Macro Retrieval: Reducing Long-Form Hallucination in Large Language Models
Apr 10, 2026Yujie Feng, Jian Li, Zhihan Zhou +7LLM Hallucination Mitigation Hallucination in Language Models
Solar System of OVB, Tencent, China · The Hong Kong Polytechnic University, Hong Kong S.A.R. · Jilin University, China
Beyond Relevance: Utility-Centric Retrieval in the LLM Era
Apr 10, 2026Hengran Zhang, Minghao Tang, Keping Bi +1Information Retrieval RAG Evaluation
State Key Laboratory of AI Safety, ICT, CAS University of Chinese Academy of Sciences Beijing, China
Guaranteeing Knowledge Integration with Joint Decoding for Retrieval-Augmented Generation
Apr 9, 2026Zhengyi Zhao, Shubo Zhang, Zezhong Wang +7Hallucination in RAG Retrieval-Augmented Generation
The Chinese University of Hong Kong · Ministry of Education Key Laboratory of High Confidence Software Technologies, CUHK · University of International Relations +2
Predicting the Benefit of Retrieval Augmentation in Open-Domain Question Answering
Apr 9, 2026Or Dado, David Carmel, Oren KurlandRetrieval-Augmented Generation Adaptive RAG
Technion, Haifa, Israel · Technology Innovation Institute (TII), Haifa, Israel
ROZA Graphs: Self-Improving Near-Deterministic RAG through Evidence-Centric Feedback
Apr 8, 2026Matthew PenarozaRetrieval-Augmented Generation Graph Retrieval-Augmented Generation
DCD: Domain-Oriented Design for Controlled Retrieval-Augmented Generation
Apr 8, 2026Valerii Kovalskii, Nikita Belov, Nikita Miteyko +2Retrieval-Augmented Generation RAG Robustness
red_mad_robot
SubSearch: Intermediate Rewards for Unsupervised Guided Reasoning in Complex Retrieval
Apr 8, 2026Roxana Petcu, Evangelos Kanoulas, Maarten de RijkeRL for Language Model Reasoning Process Reward Models
IRLab, University of Amsterdam
DTCRS: Dynamic Tree Construction for Recursive Summarization
Apr 8, 2026Guanran Luo, Zhongquan Jian, Wentao Qiu +2Text Summarization Multi-Hop QA
School of Informatics, Xiamen University
Feedback Adaptation for Retrieval-Augmented Generation
Apr 8, 2026Jihwan Bang, Seunghan Yang, Kyuhong Shim +3Retrieval-Augmented Generation Adaptive RAG
Qualcomm AI Research‡, Qualcomm Korea YH, Seoul, Republic of Korea · Sungkyunkwan University · Kyung Hee University
DQA: Diagnostic Question Answering for IT Support
Apr 7, 2026Vishaal Kapoor, Mariam Dundua, Sarthak Ahuja +6Question Answering Customer Support Automation
Amazon

Apr 16, 2026J/K move · Enter open · S save

IG-Search: Step-Level Information Gain Rewards for Search-Augmented Reasoning

Zihan Liang, Yufei Ma, Ben Chen, Zhipeng Qian+5

Kuaishou Technology

Retrieval-Augmented Generation RL for Language Model Reasoning Credit Assignment in RL Step-Level Credit Assignment LLM Reasoning

Open paper AskPDF arXiv

Reinforcement learning has emerged as an effective paradigm for training large language models to perform search-augmented reasoning. However, existing approaches rely on trajectory-level rewards that cannot distinguish precise search queries from vague or redundant ones within a rollout group, and collapse to a near-zero gradient signal whenever every sampled trajectory fails. In this paper, we propose IG-Search, a reinforcement learning framework that introduces a step-level reward based on Information Gain (IG). For each search step, IG measures how much the retrieved documents improve the model's confidence in the gold answer relative to a counterfactual baseline of random documents, thereby reflecting the effectiveness of the underlying search query. This signal is fed back to the corresponding search-query tokens via per-token advantage modulation in GRPO, enabling fine-grained, step-level credit assignment within a rollout. Unlike prior step-level methods that require either externally annotated intermediate supervision or shared environment states across trajectories, IG-Search derives its signals from the policy's own generation probabilities, requiring no intermediate annotations beyond standard question-answer pairs. Experiments on seven single-hop and multi-hop QA benchmarks demonstrate that IG-Search achieves an average EM of 0.430 with Qwen2.5-3B, outperforming the strongest trajectory-level baseline (MR-Search) by 1.6 points and the step-level method GiGPO by 0.9 points on average across benchmarks, with particularly pronounced gains on multi-hop reasoning tasks. Despite introducing a dense step-level signal, IG-Search adds only ~6.4% to per-step training wall-clock time over the trajectory-level baseline and leaves inference latency unchanged, while still providing a meaningful gradient signal even when every sampled trajectory answers incorrectly.