| Term | Meaning |
|---|---|
| LLM | Large Language Model — AI trained on text (GPT-4, Claude, Gemini) |
| prompt | The input text sent to the model |
| completion / response | The text generated by the model |
| context window | Maximum tokens the model can process at once |
| token | A unit of text (~4 characters in English) |
| temperature | Controls randomness: 0 = deterministic, 1 = creative |
| hallucination | When the model generates plausible but incorrect information |
| fine-tuning | Further training a model on domain-specific data |