거대언어모델 답변의 품질을 높이기 위한 기법 중 하나. 사용자가 입력한 원래의 프롬프트에 해당 프롬프트와 관련된 정보를 추가로 덧붙여서 “증강”한 뒤 이를 기반으로 답변을 생성하게 하기.
목차
Articles
2025-03-19 - Introducing RAG 2.0 - Contextual AI
“A typical RAG system today uses a frozen off-the-shelf model for embeddings, a vector database for retrieval, and a black-box language model for generation, stitched together through prompting or an orchestration framework. This leads to a ”Frankenstein’s monster” of generative AI: the individual components technically work, but the whole is far from optimal. These systems are brittle, lack any machine learning or specialization to the domain they are being deployed to, require extensive prompting, and suffer from cascading errors. As a result, RAG systems rarely pass the production bar.”
“The RAG 2.0 approach pretrains, fine-tunes, and aligns all components as a single integrated system, backpropagating through both the language model and the retriever to maximize performance.”
2024-11-05 - HtmlRAG : HTML is Better Than Plain Text for Modeling Retrieved Knowledge in RAG Systems
2024-09-20 - Contextual retrieval
2024-08-19 - The RAG Playbook - jxnl.co
2020 - Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks - 최초로 RAG 를 소개한 논문
Tools
See also
목차 관련 글 Don't Do RAG: When Cache-Augmented Generation is All You Need for Knowledge Tasks 이 글을 인용: “Retrieval-augmented generation (RAG) has gained traction as a powerful approach for enhancing language models b…” HtmlRAG: HTML is Better Than Plain Text for Modeling Retrieved Knowledge in RAG Systems 이 글을 인용: “Retrieval-Augmented Generation (RAG) has been shown to improve knowledge capabilities and alleviate the halluci…” A Survey on Hallucination in Large Language Models: Principles, Taxonomy, Challenges, and Open Questions 이 글을 인용: “Additionally, we delve into the current limitations faced by retrieval-augmented LLMs in combating hallucinatio…” 컨텍스트 엔지니어링 이 글을 인용: “Science because doing this right involves task descriptions and explanations, few shot examples, RAG, related (…” Thinking Beyond Tokens: From Brain-Inspired Intelligence to Cognitive Foundations for AGI and its Societal Impact 이 글을 인용: “In particular, we emphasize the rise of Agentic RAG frameworks that combine retrieval, planning, and dynamic to…” Precise Zero-Shot Dense Retrieval without Relevance Labels “While dense retrieval has been shown effective and efficient across tasks and languages, it remains difficult t…” On the Theoretical Limitations of Embedding-Based Retrieval 관련 절: Empirical Connection: Real-World Datasets Open WebUI 이 글을 인용: “It supports various LLM runners like Ollama and OpenAI-compatible APIs, with built-in inference engine for RAG,…” nose 이 글을 인용: “널리 쓰이는 중복 검사기인 jscpd는 비유하자면 RAG와 유사하다.” Survey of hallucination in natural language generation “Natural Language Generation (NLG) has improved exponentially in recent years thanks to the development of seque…” iText2KG: Incremental knowledge graphs construction using large language models “Traditional NLP methods, such as named entity recognition and relation extraction, are key in information retri…” GPTScript 이 글을 인용: “GPTScript is a framework that allows Large Language Models (LLMs) to operate and interact with various systems.” Archon (software) 이 글을 인용: “Smart search capabilities with advanced RAG strategies” Large language models are human-level prompt engineers “Reverse Mode Generation: "내 친구에게 다음과 같은 지시를 했어: <여기를 채우시오> 그랬더니 친구가 그걸 읽고 다음 질문들에 대하여 이렇게 답을 했어.” Assisting in Writing Wikipedia-like Articles From Scratch with Large Language Models “Compared to articles generated by an outline-driven retrieval-augmented baseline, more of STORM's articles are …” Automatic generation of visualizations and infographics with LLMs “In this work, we pose visualization generation as a multi-stage generation problem and argue that well-orchestr…” Programmer's brain “The best kind of practice is retrieval practice, where you try to remember information before looking it up.” Memory manipulations in extended reality “manipulations occurring pre-retrieval (i.e., just before the memory is remembered), which aim at inducing forge…”