Chain-of-thought
Prompting a model to write out intermediate reasoning before its answer, which can improve accuracy on multi-step problems.
Chain-of-thought prompting asks the model to show its working — the steps between question and answer. Because a model generates one token at a time and reads its own output as it goes, writing the steps out first gives it something to reason from, which helps on problems that genuinely have intermediate steps.
It does little for tasks with no steps to show, and newer reasoning models already do this internally, making an explicit request redundant. When you do use it, put the answer last: if the model states its conclusion first, the reasoning becomes a rationalisation rather than a cause.