AI Glossary
Plain-language definitions of the AI terms behind this product — prompts, grounding, citations, and more.
AI Mediator
A layer that turns your rough request into a clear, well-scoped prompt before the AI answers — so you get a useful reply on the first try.
Read definition →Citation
A verifiable reference that links a claim to the exact passage in a source and quotes it, so you can check any answer for yourself.
Read definition →Context Window
The maximum amount of text — prompt, history, documents and reply — a model can weigh at once, counted in tokens.
Read definition →Grounding
Grounding ties an AI's answer to specific sources you supply — your files, pasted text, or web results — instead of its fuzzy training memory.
Read definition →Hallucination
A hallucination is when an AI states something false or invented — a fake statistic, quote, or source — in the same confident tone as real fact.
Read definition →Large Language Model (LLM)
A large language model is an AI trained on vast amounts of text to predict and generate human-like language — Claude, GPT, and Gemini are examples.
Read definition →Prompt
A prompt is the instruction or question you give an AI — its wording, detail, and context strongly shape the quality of the answer you get back.
Read definition →Retrieval-Augmented Generation (RAG)
Retrieval-augmented generation fetches the most relevant source passages and hands them to the AI before it answers, so replies stay factual.
Read definition →System Prompt
Hidden instructions that set a model's role, tone, and rules for a chat — in AskClear the mediator writes them for you.
Read definition →Token
The small unit of text — roughly a word-piece — that models read and generate; in AskClear your balance is shown as money, not tokens.
Read definition →