Prompt.
The input text sent to a language model — the question, instruction, or context that triggers a response.
A prompt is any text input provided to a language model to elicit a response. Prompts can range from a simple question ('What is the capital of France?') to complex multi-part instructions with context, examples, formatting constraints, and a persona for the model to adopt. The quality and structure of a prompt dramatically affects the quality of the model's output.
Prompts typically consist of several components: a system prompt (instructions about the model's role and behavior), user messages (the actual request), and optionally few-shot examples (demonstrations of desired behavior) and context (retrieved documents, conversation history, etc.).
Prompt engineering — the practice of crafting effective prompts — has become a valuable skill. Techniques like chain-of-thought prompting, role assignment, output format specification, and self-consistency prompting can dramatically improve model outputs on complex tasks without any fine-tuning.
Related terms.
- Prompt EngineeringThe practice of crafting inputs, instructions, examples, and formatting to guide language models toward producing better, more accurate outputs.
- System PromptA special instruction given to a language model before the user conversation begins, establishing the model's persona, capabilities, constraints, and context.
- In-Context LearningThe ability of an LLM to learn from examples provided directly in the prompt, without any weight updates or training — just from seeing patterns in context.