AI Glossary

Every AI term explained at three levels — plain English, practical context, and full technical depth.

A–Z

Systems

1 terms
AgentAI Agent

An AI that can plan and take actions, not just answer questions

systemsagents

Concepts

1 terms
EmbeddingsVector Embeddings

How AI represents meaning as numbers

conceptsretrieval

Training

3 terms
Fine-tuningModel Fine-tuning

Teaching a general AI to specialize in your domain

trainingmodels
QuantizationModel Weight Quantization

Making AI models smaller without making them much dumber

traininginference
RLHFReinforcement Learning from Human Feedback

How AI learns to be helpful and safe from human ratings

trainingalignment

Infrastructure

4 terms
GGUFGGML Unified Format

The file format that lets LLMs run on any device

infrastructurequantization
MLXApple Machine Learning eXchange

Apple's ML framework built for Apple Silicon unified memory

infrastructureApple
Vector DatabaseVector Database

A database that searches by meaning, not exact words

infrastructureretrieval
vLLMVirtual LLM Serving Library

The production inference engine behind most LLM APIs

infrastructureproduction

Models

1 terms
LLMLarge Language Model

The AI brain that reads and writes text

modelsfundamentals

Protocols

1 terms
MCPModel Context Protocol

The USB standard for connecting AI to tools

protocolagents

Techniques

1 terms
Prompt EngineeringPrompt Engineering

The skill of talking to AI so it gives you better answers

techniquespractical

Architecture

2 terms
RAGRetrieval-Augmented Generation

Giving AI a searchable memory of your documents

architectureretrieval
TransformerTransformer Architecture

The invention that made modern AI possible

architecturefundamentals