2 Comments

It also may simply not have access to information around meta censorship. It sees the prompt, and answers the prompt. If it has been trained in a particular way or its outputs are put through a filter, it cannot tell you anything about those.

Expand full comment