Assess the quality, relevance and accuracy of the response and identify what is working well and what could be improved. Is the answer inaccurate? Do you want a different format? Based on the analysis, you can adjust and adapt the prompt with new instructions. This may include specific limitations and new techniques. A good deal of patience comes in handy here. With services such as Catgut, you will also be able to refer to previous replies in a conversation thread. In this way, you can also easily ask to have something elaborated, exemplified or reformulated, without having to "start over" every time.

Give a little hint The fact that you are actually having a conversation , where previous questions and answers are taken into account, means that it can sometimes be effective to give the model a hint about what it should Brazil WhatsApp Number List answer and how it should answer. This is called "priming", and can be translated into preparing or warming up the model. A prompt without priming "Explain to me what artificial intelligence is." A prompt with priming "Explain to me what artificial intelligence is. Think of computers that can learn and make decisions a bit like humans.



If you are aware of the "flow of the conversation", and perhaps also open up for follow up questions, this will help to give the language model a good context to better answer the upcoming and often more precise questions. You will find that the model sometimes follows up the conversation automatically, while at other times you have to "prime" it to remember the context and previous answers. Step by step Language models give you better explanations if they are given the opportunity to answer step by step. By adding the instruction "let's think step by step" to a prompt, the model itself can add context and explain how it arrived at the answer.