1 paper across 1 session
We present a data-centric pretraining framework that builds safety into the model from the start