I’m just curious about how to use a term to refer to all non-scalable oversight methods in outer alignment. It’s better if there is an existing term since I don’t want to reinvent new wheels.
[Question] Is there any existing term summarizing non-scalable oversight methods in outer alignment?
I’m just curious about how to use a term to refer to all non-scalable oversight methods in outer alignment. It’s better if there is an existing term since I don’t want to reinvent new wheels.