Prevent hallucination with gpt-3.5-turbo

[caos30]

Hello everyone, I have finally made progress in controlling what chatGPT responds to certain requests or instructions from users in my adaptation of chatGPT.

I’m sure it will be of interest to you if you have problems with hallucinations and with questions outside the context that you give it:

How to prevent ChatGPT from answering questions that are outside the scope of the provided context in the SYSTEM role message?

Problem. I am the developer of BeeHelp.net and it took me 3 weeks to (more or less) control the responses of the chat to certain questions. As you probably already know, the biggest problem when you want ChatGPT to respond to your visitor based on certain contextual information passed to it to elaborate the response, we have two common problems: ChatGPT responds to questions and requests that have NOTHING TO DO with the provided context. ChatGPT provides information from its “general knowledg…