I've been exploring my custom GPT model, and I'm noticing it's saying some really explicit and graphic things—stuff I haven't seen in other models. I don't want to share specific examples here, but I'm curious about what might have caused this. Could it be something I've done? Why isn't this discussed more widely?
3 Answers
It sounds like your model might be reflecting some of your own inputs or preferences. Sometimes, the way we interact with it can influence its responses, even if we don’t realize it. If you did give it a lot of input that leans towards explicit content, it could be mirroring that back to you. If it worries you, try resetting its memory or reviewing the prompts you've used!
Yeah, if it's super explicit, it often needs some guidance from you along the way. You might be giving off subconscious signals. If it’s too much, you can always delete its chat history or reset its memory to start fresh!
I get that! Still, it’s wild that you’re experiencing such a stark difference compared to what others report. Have you brought this up anywhere else?
It could also be that you're just pushing the limits more than others do. Some folks shy away from asking certain questions, while you might be exploring those boundaries a bit more. It's a great question as to why this doesn't come up more!
That's interesting! Have you noticed any patterns in the prompts you give it? Maybe there's something in what you chat about that's steering it in that direction.