Chain-of-thought is a prompting technique intended to improve the reasoning abilities of Large Language Models (LLMs). It breaks complex tasks or questions into simpler steps to support understanding and may also involve gathering additional data through web searches. A more advanced version, Tree-of-Thought, supports logical branching and backtracking.
