Chain-of-Thought
A powerful prompting strategy that forces a large language model to explicitly generate intermediate reasoning steps before outputting its final conclusion.
Think of It Like This
Like a math teacher forcing you to write out every step of your long division on the chalkboard instead of just shouting out the final number.
By breaking down complex problems into smaller, sequential logical deductions, models drastically reduce their hallucination rate on logic puzzles and math questions. This mirrors how humans use scratchpad memory to work through difficult equations rather than guessing the final answer immediately. Standard prompting just asks for the answer, while CoT asks for the step-by-step working out.