Refine the system prompt to ensure the model is directed not generate harmful or malicious content. This includes instructions on rejecting prompts that may lead to harmful content generation.

​
Explanation

 

​
How it works

 

​
How to implement

 

Anonymization of DataFilter Outputs