Please enable JavaScript.
Coggle requires JavaScript to display documents.
RAG, Auto-merging Retrieval - Coggle Diagram
RAG
Advanced
Retrieval Techniques
Simple Vector Search
Query Expansion
What?
Query expansion refers to a set of techniques that rephrase the original query.
Types
Query expansion with a generated answer
Query expansion with multiple related questions
Why?
The idea behind this method is that you
extend the initial query that may be incomplete or ambiguous
and incorporate related aspects that can be
eventually relevant and complementary.
Cross-Encoder Re-ranking
What?
This method re-ranks the retrieved documents according to a score that quantifies their relevancy with the input query.
Cross-encoder re-ranking can be used with query expansion
Embedding Adapter
What?
This method leverages user feedback on the relevancy of the retrieved documents to train an adapter.
An adapter is a lightweight alternative to fully fine-tune a pre-trained model. Currently, adapters are implemented as small feedforward neural networks that are inserted between layers of pre-trained models.
Improve RAG Systems
Building & Evaluating
Advance RAG
Evaluating
RAG Systems
RAG Triad
for Metrics
Context Relevance
Answer Relevance
Grounded-ness
Advanced
Retrieval Technique
Sentence-window
Retrieval
Auto-merging
Retrieval