Weak-to-Strong Generalization: Eliciting Strong Capabilities With Weak Supervision.
Collin BurnsPavel IzmailovJan Hendrik KirchnerBowen BakerLeo GaoLeopold AschenbrennerYining ChenAdrien EcoffetManas JoglekarJan LeikeIlya SutskeverJeff WuPublished in: CoRR (2023)