AI Common Terms
Table of Contents + −
AI tutorials use many repeated words.
This page is a simple glossary so you can revise the important terms quickly.
🤔 Why Learn These Terms?
When you read AI documentation, you will often see words such as:
- LLM
- Token
- Prompt
- Context
- Embedding
- RAG
- Agent
- Tool
- Inference
- Hallucination
If these words are unclear, AI applications feel confusing.
So let’s keep the meanings simple.
📋 Common Terms
| Term | Simple meaning |
|---|---|
| AI | Technology that enables computers to perform tasks that normally require human intelligence. |
| Machine Learning | A way of building AI systems where computers learn patterns from data. |
| Deep Learning | A type of Machine Learning that uses neural networks with multiple layers. |
| Neural Network | A model structure inspired by connected processing units called neurons. |
| Transformer | A neural-network architecture that uses attention to process relationships in a sequence. |
| LLM | Large Language Model; a model that processes and generates language. |
| Token | A small piece of text processed by a language model. |
| Tokenization | The process of breaking text into tokens. |
| Parameter | A learned numerical value inside a model. |
| Context | Information available to the model for the current request. |
| Context Window | The maximum amount of context a model can handle for a request. |
| Inference | Using a trained model to produce an output. |
| Temperature | A generation setting that affects how varied the output can be. |
| Hallucination | An incorrect or unsupported answer that sounds confident. |
| Prompt | The input or instruction given to an AI model. |
| Embedding | A vector representation created by a model to capture useful relationships. |
| RAG | Retrieval-Augmented Generation; retrieve relevant information, then generate an answer. |
| Tool | An external capability an AI application can call. |
| Agent | An AI workflow that can plan steps and use tools. |
📋 Terms That Are Easy to Confuse
Token vs Parameter
They are completely different concepts.
Context vs Training
Context is information given during the current request.
Training is the process that changes model parameters.
Prompt vs Context
A prompt is the instruction or input we give.
Context is the broader information available to the model for that request.
A prompt can be part of the context.
RAG vs Fine-Tuning
RAG gives information to the model at request time.
Fine-tuning changes model behaviour through additional training.
🧩 Key Points
- AI has many terms, but each term has a simple basic meaning.
- Tokens, parameters, context, and prompts are different things.
- RAG, tools, and agents are usually application-level ideas.
- Learning the vocabulary makes AI application architecture easier to understand.
🚀 What Comes Next?
Now we can move from fundamentals to building a simple LLM application.