Which prompt engineering technique starts with simple prompts and gradually increases complexity based on the AI's responses to guide the AI effectively while minimizing user effort?

Which prompt engineering technique starts with simple prompts and gradually increases complexity based on the AI's responses to guide the AI effectively while minimizing user effort?

• A. Least-to-Most
• B. Tree-of-Thought Prompting (ToT)
• C. Self-Consistency
• D. Generated Knowledge Prompting

Answer: (A)

Progressive prompting that starts with simple prompts and adds complexity only as needed is Least-to-Most prompting. This approach guides the AI effectively while keeping user effort low by avoiding heavy upfront instructions. Start with a very simple request, and if the model’s response signals that more guidance would help, you escalate with additional, more detailed prompts. For example, you might first ask for a brief outline, then, if the outline is insufficient, request a step-by-step plan, and finally ask for justification or specific checks. This keeps the interaction lean while ensuring the model receives the necessary scaffolding to solve the task. Tree-of-Thought prompting, in contrast, emphasizes exploring a tree of intermediate reasoning steps to improve complex problem solving, not necessarily minimizing user effort. Self-Consistency relies on sampling multiple reasoning paths and selecting the most coherent outcome, focusing on robustness of the answer rather than incremental prompting. Generated Knowledge Prompting centers on leveraging or creating external knowledge as part of the answer process, rather than progressively increasing prompt complexity.

Progressive prompting that starts with simple prompts and adds complexity only as needed is Least-to-Most prompting. This approach guides the AI effectively while keeping user effort low by avoiding heavy upfront instructions. Start with a very simple request, and if the model’s response signals that more guidance would help, you escalate with additional, more detailed prompts. For example, you might first ask for a brief outline, then, if the outline is insufficient, request a step-by-step plan, and finally ask for justification or specific checks. This keeps the interaction lean while ensuring the model receives the necessary scaffolding to solve the task.

Tree-of-Thought prompting, in contrast, emphasizes exploring a tree of intermediate reasoning steps to improve complex problem solving, not necessarily minimizing user effort. Self-Consistency relies on sampling multiple reasoning paths and selecting the most coherent outcome, focusing on robustness of the answer rather than incremental prompting. Generated Knowledge Prompting centers on leveraging or creating external knowledge as part of the answer process, rather than progressively increasing prompt complexity.