What is prompt engineering?

Since we have been discussing Amazon Bedrock models and how to invoke them, we need to dive into prompt engineering. Essentially, in a way that a particular child asks their parents questions about anything and everything, we can also ask an LLM anything under the Sun! However, to get the best and most precise outputs possible, we must train ourselves to ask the model the right questions in the right manner.

With the increasing popularity of LLMs, users are actively striving to refine their way of asking the model different kinds of questions to attain a desired response. For instance, we can simply ask an LLM questions such as Who was the first person to land on the Moon? or How many moons does Jupiter have?. Based on these questions, the language model can respond to the user’s queries either factually or provide an inadequate/incorrect response based on the LLM’s knowledge, which is the data it has been trained on.

Incorrect responses...