================================================================
AI FLUENCY — CHEAT SHEET
Prompting Techniques
================================================================

From zero-shot to chain-of-thought, ReAct, and structured outputs.
6 terms

----------------------------------------------------------------

* Chain-of-Thought (CoT)
  Prompting a model to reason step-by-step before giving a final answer, which improves accuracy on multi-step problems.
  e.g. "Let's think step by step..."

* Few-Shot Prompting
  Providing a handful of example input/output pairs in the prompt to steer the model's behavior before asking it to do the task.
  e.g. Three sample Q&A pairs followed by a new question.

* ReAct
  A prompting pattern that interleaves reasoning ("Thought") with actions ("Act") like tool calls, then observes results before continuing.
  e.g. Thought → search the web → Observation → Answer.

* Structured Output
  Constraining a model's response to a defined schema (like JSON) so downstream code can parse it reliably.
  e.g. Forcing a response to match { "name": string, "age": number }.

* System Prompt
  A privileged instruction set at the start of a conversation that shapes a model's persona, rules, and constraints for the whole session.
  e.g. "You are a concise, professional support agent."

* Zero-Shot Prompting
  Asking a model to perform a task from instructions alone, with no examples provided.
  e.g. "Translate this sentence to French."

----------------------------------------------------------------
Tip: paste this file into your favorite AI assistant and ask to be tutored on it.