Definition
A retrieval-augmentation technique that uses an LLM to generate a zero-shot pseudo-document in response to a query; this 'hypothetical' answer is then embedded and used as the vector search key to bridge the semantic gap between short user queries and long-form documents. While it improves retrieval recall, it introduces trade-offs in the form of increased latency and API costs due to the additional LLM generation step.
Not an encryption or data-masking technique; it is a 'hallucination-driven' search strategy.
"The Forensic Sketch Artist: Instead of searching a database with a vague verbal description, you create a detailed drawing of what the suspect likely looks like and search for visual matches to that drawing."
- Query Expansion(Parent Category)
- Dense Retrieval(Prerequisite)
- Embedding Model(Component)
- Zero-shot Learning(Mechanism)
Conceptual Overview
A retrieval-augmentation technique that uses an LLM to generate a zero-shot pseudo-document in response to a query; this 'hypothetical' answer is then embedded and used as the vector search key to bridge the semantic gap between short user queries and long-form documents. While it improves retrieval recall, it introduces trade-offs in the form of increased latency and API costs due to the additional LLM generation step.
Disambiguation
Not an encryption or data-masking technique; it is a 'hallucination-driven' search strategy.
Visual Analog
The Forensic Sketch Artist: Instead of searching a database with a vague verbal description, you create a detailed drawing of what the suspect likely looks like and search for visual matches to that drawing.