How would you design a language model that minimizes harmful outputs while staying useful and expressive?

candidate-reports IGotAnOffer - GenAI System Design Interview Source

01
Open

No reference answer yet

This question is in the bank but has no rubric written for it. It can be read and practised, but not scored.

—