This is part 5 of 6-part mini essays written in a simple language to help non technical users get started on Prompt Engineering
As you practice Prompt Engineering, it’s important to understand some of the more advanced concepts, shaping this landscape. You will hear more of them as you go further in your learning journey.
Fine-tuning
Adjusting an existing model on a specific task or dataset to improve the model’s performance in a particular domain. Fine-tuning can be expensive and for almost 90 to 95% of your use cases, you would not need it.
Retrieval Augmented Generation (RAG)
A popular technique to improve to ground AI output in your own data. It works by searching and then adding relevant context to a prompt. You would also hear another term ‘vector databases’ to support fast retrieval for RAG.
Hallucination
This phenomenon occurs when an AI system generates content that seems plausible but is factually incorrect or nonsensical.
Multimodal AI
Most popular LLMs are still single mode i.e. they can either generate text or image or video etc. Multi modality refers to processing and generating multiple types of data, such as text, images, and audio, often in combination.
Function Calling
Function calling is a feature that allows AI models to interact with external tools, APIs, and data sources. It enables the model to generate structured outputs that specify function names and arguments. These can then be used to call external APIs or perform specific actions. This capability enhances the model’s ability to provide relevant and contextual answers by accessing real-time information and services.
Hola 