False Prompt Blocks

I’ve noticed that words such as ‘adolescent’, ‘boy’, and their variations cause any prompt to be denied for inappropriate content, even if it is the contrary. This leads to false positives that block valid and harmless prompts.

Will there be a fix to this?

1 Like

Hi @Mr_Who_Are_You
Can you share the complete prompt and the model you are using ? Will try to reproduce from my end .