The Nonlinear Library - AF - Goals selected from learned knowledge: an alternative to RL alignment by Seth Herd
Sign in to continue reading, translating and more.