- ホーム
- > 洋書
- > 英文書
- > Computer / General
Full Description
We know that AI language models are useful, but sometimes they give answers that are not correct. This happens because they are not connected to a source of real-time facts. To build AI that you can truly trust, you need to give it a way to look up information before it gives an answer. This method is called Retrieval Augmented Generation, or RAG, and this book shows you how it works.
Learn the main ideas behind RAG in simple, easy-to-understand language.
See how RAG has grown from a simple concept into a powerful tool.
Follow step-by-step examples in Python to build your first RAG system.
Learn how to ask questions in a way that helps the AI give better answers.
Find out how to check if your RAG system is working well.
Get help choosing the right design for your own RAG projects.
Retrieval Augmented Generation, The Seminal Papers explains the important research that started it all. The original academic papers are translated into simple lessons. The book focuses on using these ideas in practice. The book gives you the tools to understand what makes RAG so effective.
After reading this book, you will know how to build AI applications that give trustworthy and fact-based answers. You can create systems that people can rely on for correct information. This book is perfect for students and developers who know the basics of Python and want to build more reliable AI.
Contents
1 HOW RAG RESEARCH PREVENTS DISASTERS
2 REVOLUTIONS IN SEMANTICS, SCALE, AND SIMILARITY
3 REALM: BIRTH OF END-TO-END TRAINABLE RAG
4 RETRIEVAL-AUGMENTED GENERATION FOR KNOWLEDGE TASKS
5 FUSION-IN-DECODER FOR MULTI-DOCUMENT PROCESSING
6 ATLAS: FEW-SHOT LEARNING WITH RETRIEVAL AUGMENTATION
7 HYDE: IMAGINING THE ANSWER BEFORE YOU SEARCH
8 RAG-FUSION: MULTI-QUERY RETRIEVAL ENHANCEMENT
9 SELF-RAG: RETRIEVAL WITH REFLECTION AND SELF-CRITIQUE
10 RAPTOR: RECURSIVE ABSTRACTIVE PROCESSING FOR TREE-ORGANIZED RETRIEVAL
11 CORRECTION, PLANNING, AND REASONING
12 GRAPHRAG
-
- 電子書籍
- 霜月記 講談社文庫
-
- 電子書籍
- 総合教育技術 2017年 1月号



