Techniques for Enhancing Prompt Effectiveness

A team of researchers is developing different methods to guide a large language model. Analyze the descriptions of their approaches below and match each approach to the most appropriate category of prompting technique.

A research lab is working on improving a language model's ability to summarize legal documents. Their process involves three phases:

1. Initially, they manually write simple, direct instructions like 'Summarize the following text.'
2. Next, they experiment with adding specific examples of good summaries to the instructions to guide the model's output style.
3. Finally, they develop an algorithm that automatically tests thousands of instruction variations to discover the most effective wording.

How do these three phases align with the standard categorization of prompting techniques?

The development of effective instructions for large language models often follows a logical progression. Arrange the following approaches in the order they are typically applied, from the most fundamental to the most advanced.