Advanced Techniques

Chunking and Indexing

• Unstructured - Pre-processing library for parsing PDFs, HTML, Word docs into clean chunks.
• Semantic Chunking - Splitting documents based on semantic similarity rather than fixed token windows.

Evaluation

• RAGAS - free evaluation framework. Metrics: faithfulness, answer relevancy, context precision/recall. |
• TruLens - specific metrics. |
• DeepEval - aware metrics. |
• ARES

Query Transformation

• Step-Back Prompting - Ask a more abstract question first to retrieve broader context.

Reranking

• Cohere Rerank - Cross-encoder reranking API.
• ColBERT - Late interaction model for efficient and effective reranking.
• bge-reranker - Open-source cross-encoder reranker by BAAI.
• RankLLM - Using LLMs themselves as rerankers via listwise prompting.

Retrieval Strategies

• Hypothetical Document Embeddings - Generate a hypothetical answer first, then use it as the retrieval query. |
• Anthropic's approach - Prepend chunk-specific context before embedding to reduce retrieval failures. |

Datasets and Benchmarks

Evaluation

• HotpotQA - hop QA requiring reasoning over multiple documents. |
• MS MARCO - scale passage retrieval and QA benchmark. |
• BEIR - shot evaluation of retrieval models across diverse tasks. |
• RGB (Retrieval-Augmented Generation Benchmark)

Embedding Models

Evaluation

• text-embedding-3-small/large - purpose embeddings. large variant has 3072 dimensions. |
• Cohere Embed v3
• BGE (BAAI) - source | Top-performing open-source embeddings. Available in multiple sizes. |
• E5-Mistral - source | LLM-based embedding model. Strong performance on MTEB benchmark. |
• Nomic Embed - source | Long-context (8192 tokens), fully open-source with open training data. |
• Jina Embeddings - source | Multilingual, supports 8192 token context. Good for non-English corpora. |
• MTEB Leaderboard - to-date embedding model benchmarks.

Foundational Papers

• Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks - The original RAG paper by Lewis et al. (Meta AI). Introduces the RAG architecture combining a pre-trained seq2seq model with a dense retriever (DPR).
• Dense Passage Retrieval for Open-Domain Question Answering - DPR — the dense retrieval method that underpins many RAG systems.
• REALM: Retrieval-Augmented Language Model Pre-Training - Pre-trains a language model jointly with a knowledge retriever.
• Attention Is All You Need - The Transformer architecture — foundational to all modern LLMs used in RAG.

Frameworks and Libraries

Evaluation

• LangChain
• LlamaIndex
• Haystack - ready NLP framework by deepset. Pipeline-based architecture. |
• RAGFlow - source RAG engine with deep document understanding and chunk visualization. |
• Verba - source RAG chatbot powered by Weaviate. Good for quick prototyping. |
• Cognita - source modular RAG framework for production use. |

Survey Papers

• Retrieval-Augmented Generation for Large Language Models: A Survey - Comprehensive survey covering Naive RAG, Advanced RAG, and Modular RAG paradigms. Excellent starting point.
• A Survey on RAG Meeting LLMs: Towards Retrieval-Augmented Large Language Models - Covers the evolution of RA-LLMs, taxonomies, and training strategies.
• Seven Failure Points When Engineering a RAG System - Practical guide to what can go wrong in RAG pipelines — highly recommended for thesis work.

Tutorials and Guides

Evaluation

• RAG From Scratch (LangChain) - Series of notebooks covering RAG concepts from basics to advanced patterns.
• Building RAG Applications with LlamaIndex - Official LlamaIndex documentation and conceptual guide.
• Pinecone RAG Learning Center - Well-written introduction to RAG with practical examples.
• Full Stack RAG App Tutorial (freeCodeCamp) - Video walkthrough of building a complete RAG application.
• Building RAG Applications with LlamaIndex - Official LlamaIndex documentation and conceptual guide.

Vector Databases

Evaluation

• Chroma
• Weaviate - hosted / Cloud | Supports hybrid search natively. GraphQL API. |
• Qdrant - hosted / Cloud | Written in Rust. Excellent filtering and payload support. |
• Milvus - hosted / Cloud | Highly scalable. Used in many production deployments. |
• Pinecone - only | Fully managed. Simple API. Popular in industry. |
• FAISS
• pgvector

Videos and Talks

Evaluation

• But what is RAG? (3Blue1Brown-style explainer) - Visual, intuitive explanation of how RAG works.
• RAG is Dead? Long Live RAG! (Keynote) - Discussion on the future of RAG vs. long-context models.
• Building Production RAG (AI Engineer Summit) - Practical lessons from deploying RAG at scale.
• Advanced RAG Techniques (DeepLearning.AI) - Short course by Andrew Ng's platform.