Hi there,
When I test my chatbot designed for university student service, i found a problem affected the accuracy level of answers.
There is difference between when I send question after lots of questions and when I start run the chatbot and use this question as my first question. The former cannot be identified, and the latter can be identified. I have no idea how to solve this unless refresh. but I can't expect users refresh bot once ask a new question.
Thanks in advance!
Good morning,
I hope you are doing well. This behaviour can come from multiple factors : generative AI keep past questions in memory and can become stuck when conversations get long.
In order to avoid this, I would redirect to the reset topic each time a generative AI question was asked.
Have a great day!