AI Dictionary

AI Dictionary is a hands-on introduction to artificial intelligence using annotated definitions and examples. Check out the first definition or browse the full list below.

Agent
Embedding
Fine-tuning
Hallucination
LLM
Prompt Engineering
RAG
Temperature
Token
Vector Database

AI Dictionary: Agent

In AI, an agent is an autonomous system that perceives its environment, makes decisions, and takes actions to achieve specific goals.

Modern AI agents can use tools, call APIs, and interact with external systems independently. They combine perception, reasoning, and action in a goal-directed manner.

autonomous tool-calling goal-oriented
agent.perceive(environment)
decision = agent.reason(perception)
result = agent.act(decision)
// Autonomous goal achievement

Next definition: Embedding

AI Dictionary: Embedding

An embedding is a numerical representation of text, images, or other data as vectors in high-dimensional space.

Similar concepts have similar embeddings, enabling semantic search and similarity matching. They're the foundation of modern AI search systems.

vector representation semantic similarity
text = "artificial intelligence"
embedding = model.encode(text)
// [0.1, -0.3, 0.8, ..., 0.4]
// 1536-dimensional vector

Next definition: Fine-tuning

AI Dictionary: RAG

RAG (Retrieval-Augmented Generation) combines information retrieval with text generation, allowing models to access external knowledge.

Essential for creating AI systems that can answer questions about specific documents or data that wasn't in the model's training set.

external knowledge document Q&A
docs = retrieve(query, knowledge_base)
context = docs + query
answer = llm.generate(context)
// Grounded in retrieved knowledge

Next definition: Temperature

AI Dictionary: Token

A token is the basic unit of text that AI models process—roughly 3/4 of a word in English.

Tokens determine context limits and pricing for most AI APIs. Understanding tokens is crucial for cost management and system design.

text unit context limits API pricing
"Hello world" → ["Hello", " world"]
"understanding" → ["under", "standing"]
cost = tokens × price_per_token
// ~4 tokens per 3 words

Next definition: Vector Database