RAG (Retrieval-Augmented Generation)
Combining retrieval systems with language models to generate responses grounded in external knowledge.
In-depth explanation
RAG addresses LLM limitations like outdated knowledge and hallucinations by retrieving relevant documents before generation. The retrieved context is included in the prompt, grounding the response in actual sources. This enables knowledge-intensive applications, reduces hallucinations, and allows updating knowledge without retraining the model.
Examples
More in Generative AI
GPT
Generative Pre-trained Transformer, a family of large language models trained to generate text.
Large Language Model (LLM)
AI models trained on vast text data that can generate and understand human-like text.
Prompt Engineering
The practice of crafting effective inputs to get desired outputs from AI models.
Generative Adversarial Network (GAN)
Two neural networks competing against each other to generate realistic synthetic data.
Diffusion Model
Generative models that learn to create data by reversing a gradual noising process.
Hallucination
When AI models generate plausible-sounding but factually incorrect or fabricated information.
Master RAG (Retrieval-Augmented Generation).
Learn how to apply this concept with hands-on projects in our comprehensive AI programs.