AI 요약
We present a new research direction for superalignment, together with promising initial results: can we leverage the generalization properties of deep learning to control strong models with weak supervisors?
댓글0개
댓글을 작성하려면 로그인이 필요합니다.
Weak-to-strong generalization
We present a new research direction for superalignment, together with promising initial results: can we leverage the generalization properties of deep learning to control strong models with weak supervisors?