Hey all, has anyone found a way to prevent the bot...
# đź’»developers
r
Hey all, has anyone found a way to prevent the bot (or GPT-3.5 really) from appending the "please note...", "for more info..." type of statements to the KB answer? Currently I'm feeding its original answer into another AI task that removes those statements, and although it works pretty well, it often adds a several seconds to the response time. I have tried putting those instructions in the KB agent but it ignores them. I also tried the personality agent, but that rephrased every answer, often in ways that were less desirable.