Learn Before
Example

Illustrating the Need for Decomposition in Generative Tasks

When assigning a complex generative task to a Large Language Model, such as writing a blog about AI risks, a simple, high-level prompt like 'write a blog about the risks of AI' can result in an output with an arbitrary and potentially disorganized structure. This example highlights the advantage of decomposing the problem first, for instance by creating an outline, to guide the LLM in producing a more structured and coherent result.

0

1

Updated 2025-10-10

Contributors are:

Who are from:

Tags

Ch.3 Prompting - Foundations of Large Language Models

Foundations of Large Language Models

Foundations of Large Language Models Course

Computing Sciences

Related