The Chain-of-Thought Technique

Improve complex reasoning by asking the model to 'think step by step' before giving its final answer. This simple instruction prompts LLMs to break down the problem-solving process, often resulting in more accurate answers.