发帖
雷达卡
本仓库依托综述论文《Retrieval-Augmented Generation for AI-Generated Content: A Survey》整理而成，用于检索增强生成（RAG） 相关论文的收集与分类：Retrieval-Augmented Generation for AI-Generated Content: A Survey。鉴于该领域发展迅速，我们将持续更新本篇综述论文与本仓库内容。
Overview 整体概览
Catalogue 目录体系
Methods Taxonomy 方法体系分类
RAG Foundations RAG 基础架构
Query-based RAG 基于查询的检索增强生成
REALM: Retrieval-Augmented Language Model Pre-Training REALM：面向语言模型预训练的检索增强方法 Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflection Self-RAG：基于自反思的检索、生成与纠错一体化模型 REPLUG: Retrieval-Augmented Black-Box Language Models REPLUG：面向黑盒语言模型的检索增强框架 In-Context Retrieval-Augmented Language Models 上下文检索增强语言模型 When Language Model Meets Private Library 语言模型与私有知识库的融合方案 DocPrompting: Generating Code by Retrieving the Docs 文档提示：基于文档检索的代码生成方法 Retrieval-based prompt selection for code-related few-shot learning 面向代码小样本学习的检索式提示词筛选 Inferfix: End-to-end program repair with llms Inferfix：基于大语言模型的端到端程序修复 Make-an-audio: Text-to-audio generation with prompt-enhanced diffusion models Make-an-audio：结合提示增强扩散模型的文生音频 Reacc: A retrieval-augmented code completion fr amework Reacc：检索增强代码补全框架 Uni-parser: Unified semantic parser for question answering on knowledge base and database Uni-parser：面向知识库与数据库问答的统一语义解析器 RNG-KBQA: generation augmented iterative ranking for knowledge base question answering RNG-KBQA：融合生成与迭代排序的知识库问答方法 End-to-end casebased reasoning for commonsense knowledge base completion 面向常识知识库补全的端到端案例推理 Combining transfer learning with in-context learning using blackbox llms for zero-shot knowledge base question answering 结合迁移学习与上下文学习的黑盒大模型零样本知识库问答 Genegpt: Augmenting large language models with domain tools for improved access to biomedical information GeneGPT：融合领域工具增强大模型，优化生物医学信息检索 Retrieval-augmented large language models for adolescent idiopathic scoliosis patients in shared decision-making 面向青少年特发性脊柱侧弯协同决策的检索增强大语言模型 Retrievegan:Image synthesis via differentiable patch retrieval RetrieveGAN：基于可微图像块检索的图像合成模型 Instance-conditioned gan 实例条件生成对抗网络 Retrieval-Augmented Score Distillation for Text-to-3D Generation 面向文生三维内容的检索增强分数蒸馏方法
Latent Representation-based RAG 基于隐层表征的检索增强生成
Leveraging passage retrieval with generative models for open domain question answering 融合段落检索与生成模型的开放域问答 Bashexplainer: Retrieval-augmented bash code comment generation based on finetuned codebert BashExplainer：基于微调 CodeBERT 的检索增强 Bash 代码注释生成 EditSum: A Retrieve-and-Edit fr amework for Source Code Summarization EditSum：面向代码摘要的检索-编辑框架 Retrieve and Refine: Exemplar-based Neural Comment Generation 检索与优化：基于范例的神经代码注释生成 RACE: retrieval-augmented commit message generation RACE：检索增强代码提交信息生成 Unik-qa: Unified representations of structured and unstructured knowledge for open-domain question answering Unik-QA：融合结构化与非结构化知识表征的开放域问答 A Retrieve-and-Edit fr amework for Predicting Structured Outputs 面向结构化输出预测的检索-编辑框架 DecAF: Joint Decoding of Answers and Logical Forms for Question Answering over Knowledge Bases DecAF：知识库问答中答案与逻辑表达式联合解码方法 Bridging the kb-text gap: Leveraging structured knowledge-aware pre-training for KBQA 弥合知识库与文本鸿沟：面向知识库问答的结构化知识预训练 Knowledge-driven cot: Exploring faithful reasoning in llms for knowledge-intensive question answering 知识驱动思维链：知识密集型问答中大模型的可信推理研究 Retrieval-enhanced generative model for large-scale knowledge graph completion 面向大规模知识图谱补全的检索增强生成模型 Case-based reasoning for natural language queries over knowledge bases 面向知识库自然语言查询的案例推理 A Protein-Ligand Interaction-focused 3D Molecular Generative fr amework for Generalizable Structure-based Drug Design 面向蛋白-配体相互作用、具备泛化能力的三维分子生成药物设计框架 Improving language models by retrieving from trillions of tokens 基于万亿级词元检索优化语言模型性能 Remodiffuse: Retrieval-augmented motion diffusion model ReMoDiffuse：检索增强动作扩散模型 Memorizing transformers 记忆型 Transformer 模型 Audio captioning using pre-trained large-scale language model guided by audio-based similar caption retrieval 结合相似音频描述检索与预训练大语言模型的音频字幕生成 Retrieval augmented convolutional encoder-decoder networks for video captioning 检索增强卷积编解码网络的视频字幕生成 Retrieval-augmented egocentric video captioning 检索增强第一视角视频字幕生成 Re-imagen: Retrievalaugmented text-to-image generator Re-Imagen：检索增强文生图模型 Knn-diffusion: Image generation via large-scale retrieval kNN-Diffusion：基于大规模检索的图像生成 Retrieval-augmented diffusion models 检索增强扩散模型 Text-guided synthesis of artistic images with retrieval-augmented diffusion models 基于检索增强扩散模型的文本引导艺术图像生成 Memory-driven text-to-image generation 记忆驱动型文生图 Mention memory: incorporating textual knowledge into transformers through entity mention attention 提及记忆：借助实体提及注意力将文本知识融入 Transformer Unlimiformer:Long-range transformers with unlimited length input Unlimiformer：支持超长输入的长文本 Transformer Entities as experts: Sparse memory access with entity supervision 实体专家机制：基于实体监督的稀疏内存访问 Amd: Anatomical motion diffusion with interpretable motion decomposition and fusion AMD：具备可解释动作分解与融合的人体动作扩散模型 Retrieval-augmented text-to-audio generation 检索增强文生音频 Concept-aware video captioning: Describing videos with effective prior information 概念感知型视频字幕生成：利用有效先验信息描述视频内容
Logit-based RAG 基于对数概率的检索增强生成
Generalization through memorization: Nearest neighbor language models 基于记忆实现泛化：近邻语言模型 Syntax-Aware Retrieval Augmented Code Generation 语法感知型检索增强代码生成 Memory-augmented image captioning 记忆增强图像字幕生成 Retrieval-based neural source code summarization 基于检索的神经代码摘要生成 Efficient nearest neighbor language models 高效近邻语言模型 Nonparametric masked language modeling 非参数掩码语言建模 Editsum:A retrieve-and-edit fr amework for source code summarization EditSum：面向代码摘要的检索-编辑框架
Speculative RAG 推测式检索增强生成
REST: Retrieval-Based Speculative Decoding REST：基于检索的推测解码 GPTCache GPT 缓存系统 COPY IS ALL YOU NEED 复制机制核心解码方法 RETRIEVAL IS ACCURATE GENERATION 检索即精准生成
RAG Enhancements RAG 优化增强技术
Input Enhancement 输入层增强
Query Transformations 查询优化
Query2doc: Query Expansion with Large Language Models Query2Doc：基于大语言模型的查询扩展 Tree of Clarifications: Answering Ambiguous Questions with Retrieval-Augmented Large Language Models 澄清树：基于检索增强大模型的歧义问题解答 Precise Zero-Shot Dense Retrieval without Relevance Labels 无相关标注的精准零样本稠密检索 RQ-RAG: Learning to Refine Queries for Retrieval Augmented Generation RQ-RAG：面向检索增强生成的查询优化学习 Dynamic Contexts for Generating Suggestion Questions in RAG Based Conversational Systems 检索增强对话系统中推荐问题生成的动态上下文构建
Data Augmentation 数据增强
LESS: selecting influential data for targeted instruction tuning LESS：面向定向指令微调的高价值数据筛选 Make-An-Audio: Text-To-Audio Generation with Prompt-Enhanced Diffusion Models Make-an-audio：提示增强扩散模型实现文生音频 Telco-RAG: Navigating the challenges of retrieval-augmented language models for telecommunications Telco-RAG：面向电信领域的检索增强语言模型落地方案
Retriever Enhancement 检索器增强
Recursive Retrieve 递归检索
Query Expansion by Prompting Large Language Models 基于大模型提示的查询扩展 Rat: Retrieval augmented thoughts elicit context-aware reasoning in long-horizon generation RAT：检索增强思维链实现长序列生成的上下文感知推理 React: Synergizing reasoning and acting in language models ReAct：语言模型推理与行动协同框架 Chain-of-thought prompting elicits reasoning in large language models 思维链提示：激发大语言模型推理能力 Large Language Models Know Your Contextual Search Intent: A Prompting fr amework for Conversational Search 大模型感知上下文检索意图：面向对话式搜索的提示框架 ACTIVERAG: Revealing the Treasures of Knowledge via Active Learning ActiveRAG：基于主动学习挖掘知识价值 Retrieval-Augmented Thought Process as Sequential Decision Making 将检索增强思维过程建模为序列决策问题 In search of needles in a 10m haystack: Recurrent memory finds what llms miss 海量数据精准检索：循环记忆弥补大模型检索短板 Lost in the middle: How language models use long contexts 中间位置遗忘问题：语言模型对长上下文的利用特性
Chunk Optimization 文本分块优化
LlamaIndex LlamaIndex：检索增强开源框架 RAPTOR: RECURSIVE ABSTRACTIVE PROCESSING FOR TREE-ORGANIZED RETRIEVAL RAPTOR：面向树形检索的递归摘要处理 Prompt-RAG: Pioneering Vector Embedding-Free Retrieval-Augmented Generation in Niche Domains, Exemplified by Korean Medicine Prompt-RAG：小众领域无向量嵌入检索增强生成（以韩医学为例） Question-Based Retrieval using Atomic Units for Enterprise RAG 面向企业级 RAG 的原子单元式问题检索
Finetune Retriever 检索器微调
C-Pack: Packaged Resources To Advance General Chinese Embedding C-Pack：面向通用中文嵌入的开源资源包 BGE M3-Embedding: Multi-Lingual, Multi-Functionality, Multi-Granularity Text Embeddings Through Self-Knowledge Distillation BGE M3-Embedding：基于自知识蒸馏的多语言、多功能、多粒度文本嵌入模型 LM-Cocktail: Resilient Tuning of Language Models via Model Merging LM-Cocktail：基于模型融合的语言模型稳健微调 Retrieve Anything To Augment Large Language Models 通用检索增强大语言模型框架 Replug: Retrieval-augmented black-box language models Replug：黑盒语言模型检索增强方案 When Language Model Meets Private Library 语言模型与私有知识库融合 EditSum: A Retrieve-and-Edit fr amework for Source Code Summarization EditSum：代码摘要检索-编辑框架 Synchromesh: Reliable Code Generation from Pre-trained Language Models Synchromesh：基于预训练模型的高可靠代码生成 Retrieval Augmented Convolutional Encoder-decoder Networks for Video Captioning 检索增强卷积编解码网络实现视频字幕生成 Reinforcement learning for optimizing RAG for domain chatbots 基于强化学习优化领域对话机器人的检索增强生成
Hybrid Retrieve 混合检索
RAP-Gen: Retrieval-Augmented Patch Generation with CodeT5 for Automatic Program Repair RAP-Gen：结合 CodeT5 与检索增强补丁生成的自动程序修复 ReACC: A Retrieval-Augmented Code Completion fr amework ReACC：检索增强代码补全框架 Retrieval-based neural source code summarization 基于检索的神经代码摘要 BashExplainer: Retrieval-Augmented Bash Code Comment Generation based on Fine-tuned CodeBERT BashExplainer：基于微调 CodeBERT 的检索增强 Bash 注释生成 Retrieval-Augmented Score Distillation for Text-to-3D Generation 文生三维内容的检索增强分数蒸馏 Corrective Retrieval Augmented Generation 纠错式检索增强生成 Retrieval augmented generation with rich answer encoding 融合丰富答案编码的检索增强生成 Unims-rag: A unified multi-source retrieval-augmented generation for personalized dialogue systems Unims-RAG：面向个性化对话系统的统一多源检索增强生成 You'll Never Walk Alone: A Sketch and Text Duet for Fine-Grained Image Retrieval 草图与文本协同的细粒度图像检索 Blended RAG: Improving RAG (Retriever-Augmented Generation) Accuracy with Semantic Search and Hybrid Query-Based Retrievers Blended RAG：结合语义检索与混合查询检索器提升 RAG 精度
Re-ranking 重排序
Re2G: Retrieve, Rerank, Generate Re2G：检索-重排-生成流水线 Passage Re-ranking with BERT 基于 BERT 的段落重排序 AceCoder: Utilizing Existing Code to Enhance Code Generation AceCoder：利用已有代码增强代码生成能力 XRICL: Cross-lingual Retrieval-Augmented In-Context Learning for Cross-lingual Text-to-SQL Semantic Parsing XRICL：面向跨语言 Text-to-SQL 语义解析的跨语言检索增强上下文学习 A Fine-tuning Enhanced RAG System with Quantized Influence Measure as AI Judge 结合量化影响评估与 AI 评判的微调增强 RAG 系统 UDAPDR: Unsupervised Domain Adaptation via LLM Prompting and Distillation of Rerankers UDAPDR：基于大模型提示与重排器蒸馏的无监督领域自适应 Learning to Retrieve In-Context Examples for Large Language Models 面向大语言模型的上下文样例检索学习 The Chronicles of RAG: The Retriever, the Chunk and the Generator RAG 技术全解析：检索器、文本分块与生成器 Enhancing LLM Factual Accuracy with RAG to Counter Hallucinations: A Case Study on Domain-Specific Queries in Private Knowledge-Bases 借助 RAG 提升大模型事实准确率、缓解幻觉：私有知识库领域查询案例
Retrieval Transformation 检索变换
Learning to filter context for retrieval-augmented generation 面向检索增强生成的上下文过滤学习 Fid-light: Efficient and effective retrieval-augmented text generation FID-Light：轻量高效的检索增强文本生成 Gar-meets-rag paradigm for zero-shot information retrieval GAR-RAG 融合范式实现零样本信息检索
Others 其他
PineCone PineCone：向量数据库 Generate rather than retrieve: Large language models are strong context generators 生成替代检索：大语言模型作为高性能上下文生成器 Generator-retriever-generator: A novel approach to open-domain question answering 生成-检索-生成：新型开放域问答框架 Multi-Head RAG: Solving Multi-Aspect Problems with LLMs 多头 RAG：基于大模型解决多维度复杂问题
Generator Enhancement 生成器增强
Prompt Engineering 提示工程
Prompt Engineering Guide 提示工程指南 Take a Step Back: Evoking Reasoning via Abstraction in Large Language Models 回溯思考：借助抽象思维激发大模型推理能力 Active Prompting with Chain-of-Thought for Large Language Models 结合思维链的大模型主动提示 Chain-of-Thought Prompting Elicits Reasoning in Large Language Models 思维链提示激发大模型推理 LLMLingua: Compressing Prompts for Accelerated Inference of Large Language Models LLMLingua：提示词压缩以加速大模型推理 Lost in the Middle: How Language Models Use Long Contexts 中间遗忘：语言模型的长上下文使用特性 ReMoDiffuse: Retrieval-Augmented Motion Diffusion Model ReMoDiffuse：检索增强动作扩散模型 Automatic Semantic Augmentation of Language Model Prompts (for Code Summarization) 面向代码摘要的大模型提示词自动语义增强 Retrieval-Based Prompt Selection for Code-Related Few-Shot Learning 代码小样本学习的检索式提示词筛选 XRICL: Cross-lingual Retrieval-Augmented In-Context Learning for Cross-lingual Text-to-SQL Semantic Parsing 跨语言检索增强上下文学习用于跨语言 Text-to-SQL 解析 Make-An-Audio: Text-To-Audio Generation with Prompt-Enhanced Diffusion Models 提示增强扩散模型实现文生音频
Decoding Tuning 解码策略优化
InferFix: End-to-End Program Repair with LLMs InferFix：基于大模型的端到端程序修复 Synchromesh: Reliable Code Generation from Pre-trained Language Models Synchromesh：预训练模型驱动的高可靠代码生成
Finetune Generator 生成器微调
Improving Language Models by Retrieving from Trillions of Tokens 万亿级词元检索优化语言模型 When Language Model Meets Private Library 语言模型与私有知识库融合 CodeGen: An Open Large Language Model for Code with Multi-Turn Program Synthesis CodeGen：面向多轮程序合成的开源代码大语言模型 Concept-Aware Video Captioning: Describing Videos With Effective Prior Information 概念感知视频字幕生成 Animate-A-Story: Storytelling with Retrieval-Augmented Video Generation Animate-A-Story：结合检索增强视频生成的故事创作 Lora: Low-rank adaptation of large language models LoRA：大语言模型低秩适配 Retrieval-Augmented Score Distillation for Text-to-3D Generation 文生三维内容的检索增强分数蒸馏
Result Enhancement 结果层增强
Rewrite Output 输出重写
Automated Code Editing with Search-Generate-Modify 检索-生成-修改流水线实现自动化代码编辑 Repair Is Nearly Generation: Multilingual Program Repair with LLMs 修复即生成：基于大模型的多语言程序修复 Case-based Reasoning for Natural Language Queries over Knowledge Bases 知识库自然语言查询的案例推理
RAG Pipeline Enhancement RAG 全流水线增强
Adaptive Retrieval 自适应检索
Rule-Based 基于规则
Active retrieval augmented generation 主动式检索增强生成 Efficient Nearest Neighbor Language Models 高效近邻语言模型 Generalization through Memorization: Nearest Neighbor Language Models 基于记忆泛化的近邻语言模型 Nonparametric masked language modeling 非参数掩码语言建模 When Not to Trust Language Models: Investigating Effectiveness of Parametric and Non-Parametric Memories 大模型可信度研判：参数化与非参数化记忆效果对比 How Can We Know When Language Models Know? On the Calibration of Language Models for Question Answering 大模型知识可信度校准：问答场景应用 Large Language Models Struggle to Learn Long-Tail Knowledge 大模型长尾知识学习难点研究
Model-Based 基于模型
Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflection Self-RAG：自反思式检索、生成、纠错一体化模型 Investigating the Factual Knowledge Boundary of Large Language Models with Retrieval Augmentation 借助检索增强探究大模型的事实知识边界 Self-Knowledge Guided Retrieval Augmentation for Large Language Models 基于自知识引导的大模型检索增强 Retrieve only when it needs: Adaptive retrieval augmentation for hallucination mitigation in large language models 按需检索：缓解大模型幻觉的自适应检索增强 Adaptive-rag: Learning to adapt retrieval-augmented large language models through question complexity Adaptive-RAG：基于问题复杂度自适应调整检索增强大模型
Iterative RAG 迭代式 RAG
RepoCoder: Repository-Level Through Iterative Retrieval and Generation RepoCoder：基于迭代检索与生成的代码仓库级代码生成 Enhancing Retrieval-Augmented Large Language Models with Iterative Retrieval-Generation Synergy 基于检索-生成协同迭代优化检索增强大模型 Knowledge graph based synthetic corpus generation for knowledge-enhanced language model pre-training 基于知识图谱的语料生成，用于知识增强语言模型预训练
推荐学习书籍 《CDA一级教材》适合CDA一级考生备考，也适合业务及数据分析岗位的从业者提升自我。完整电子版已上线CDA网校，累计已有10万+在读~ !
京公网安备 11010802022788号
