Retrieval-Augmented Generation
AI architecture that combines generative models with external search or databases to ground responses in retrieved information.
Core metadata
- ID: retrieval_augmented_generation
- Era: Modern
- First known date: 2020 (exact)
- Region: Global AI research community
- Review status: source_checked
- Maturity: established
Prerequisites
Dependents
- AI Personal Memory Systems (ai_personal_memory_systems)
- Tool-Using Language Models (tool_using_language_models)
Fields
Field lanes
- Artificial Intelligence & Machine Learning: Foundation Models
Node sources
- Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks (NeurIPS, 2020, primary_paper) • Supports: node, maturity
Prerequisite edge evidence
Edge/source evidence summary:
- Prerequisite edges: 2
- Average edge confidence: 77%
- Prerequisite sources: 2
- expert_inference: 1
- primary_source: 1
| Prerequisite | Type | Confidence | Evidence level | Note | Sources |
|---|---|---|---|---|---|
| Large Language Models (large_language_models) | enabling | 68% | expert_inference | Large Language Models provides a capability that enables this technology without being the only possible path. |
|
| Search Engines (search_engines) | enabling | 85% | primary_source | RAG requires retrieval from an external corpus, but the retrieval system can be classic search or another database, not necessarily a vector database. |
|
This page is generated from canonical era JSON and is indexable by URL.