Learn Before
Retrieval-Augmented Generation (RAG) as an Application of Reference Information
Retrieval-Augmented Generation (RAG) is a specific application of providing reference information in prompts. In this approach, an Information Retrieval (IR) system is first used to fetch text relevant to a user's query. Subsequently, a Large Language Model is prompted to generate a response based on the retrieved text, confining its answer to the provided information.
0
1
Tags
Ch.3 Prompting - Foundations of Large Language Models
Foundations of Large Language Models
Foundations of Large Language Models Course
Computing Sciences
Related
Constraining LLM Outputs with Provided Text
Leveraging Prior Knowledge in Prompts for Real-World Problems
A user wants a large language model to answer questions about the internal policies of a specific, private company. The model was not trained on this company's private data. Which of the following prompting strategies would be most effective for ensuring the model provides accurate, relevant answers based on the company's actual policies?
Customer Support Chatbot Prompt Design
Retrieval-Augmented Generation (RAG) as an Application of Reference Information
A developer is creating a feature to summarize newly published, highly technical research papers for a general audience. The language model being used has a knowledge cut-off from two years ago. To ensure the summaries are accurate and reflect the content of the new papers, the developer includes the full text of each paper within the prompt before asking for a summary. What is the primary analytical reason this approach is effective?
Learn After
Factual Accuracy for a Corporate Chatbot
A company implements a chatbot to answer employee questions based on a private, internal knowledge base of company policies. An employee asks the chatbot, 'What is the company's policy on parental leave?' The chatbot provides a detailed, accurate, and up-to-date answer. The employee then asks, 'What are the best public parks near the main office?' The chatbot responds, 'Based on the provided documents, I cannot find any information about public parks.' Which statement best analyzes the chatbot's behavior in the second interaction?
A system is designed to answer a user's question using a large, private collection of documents. The system first finds relevant information within the documents and then uses that information to construct a natural language answer. Arrange the following steps to reflect the correct operational sequence of this system.
Prompt Template for Answer Generation from Context