Myungjin Choi

RAG From Scratch: Query Translation (HyDE) Our RAG From Scratch video series walks through impt RAG concepts in short / focused videos w/ code. This is our final video on Query Translation, focused on Hypothetical Document Embeddings (HyDE) from Gao et al. 🔧 Problem: Documents are often very different from input user questions. But, many retrieval approaches embed documents and question into the same embedding space for similarity search. 💡Idea: Use an LLM to convert questions into hypothetical documents that answer the question. Use the embedded hypothetical documents to retrieve real documents with the premise that doc-doc similarity search can produce more relevant matches. 📽️ Video: youtube.com/watch?v=SaDzIVkY… 💻 Code: github.com/langchain-ai/rag-… 🧠 References: HyDE: arxiv.org/abs/2212.10496