LW - Does reducing the amount of RL for a given capability level make AI safer? by Chris Leong | The Nonlinear Library | Podwise