08 Revision v2

When adjusting parameters, it is effective to specify:

• A few examples
• The length of the response
• Multiple formats simultaneously
• The AI's processing speed

An example of providing examples in a prompt would be:

• "Write something."
• "Avoid details."
• "Just do what you think is best."
• "Write a short story title, like Alice in Wonderland."

A good practice in formatting responses is to:

• Specify whether you want a list, essay, or summary
• Use the most complex format available
• Ask for different formats in one prompt
• Leave the format entirely to the AI

Chaining AIs is effective for tasks that:

• Are too simple for a single prompt
• Need multiple specialized steps or perspectives
• Can be completed quickly by one prompt
• Require a single step

The best way to give direction in a prompt is to:

• Include multiple requests in one prompt
• Define scope and specific to the point
• Use a conversational tone
• Focus on unrelated details

In what scenario might zero-shot prompting be less effective compared to few-shots prompting?

• When the user has no idea to provide examples
• When the task is extremely simple
• When the task requires creative thinking
• When the response has a specific pattern

How does the choice of the example in one-shot prompting impact ChatGPT's output?

• A well-chosen example can significantly shape the model's response
• The example has no impact on the output
• The example only affects the length of the response
• A more complex example always yields a better response

What is a key advantage of few-shots prompting over one-shot prompting?

• It provides a more accurate and contextually relevant outputs.
• Few-shots prompting is more suitable for mathematical problems
• It ensures shorter responses
• Few-shots prompting is faster

In role-based prompting, how does the specificity of the role affect ChatGPT's responses?

• More specific roles limit the creativity of responses
• Less specific roles are always preferable for accurate information
• More specific roles enhance the accuracy of ChatGPT in context-specific scenarios
• The specificity of the role has no impact on the responses

Chain of thought prompting can improve the accuracy of responses by encouraging step-by-step reasoning.

• False
• True