What is self-ask prompting?

Self-ask prompting is a technique used with large language models (LLMs), where the model formulates additional clarifying questions about the original query to gradually obtain a more accurate and complete answer. Instead of answering immediately, the model breaks down the task into sub-tasks, asking itself questions and answering them in turn. This helps improve the quality and justification of the answers, especially when solving complex or multi-step problems.