Jojo Yang

Aug 22, 2024 | lesswrong.com | Winnie Yang |Jojo Yang |Lennart Buerger

As models grow increasingly sophisticated, they will surpass human expertise. It is a fundamentally difficult challenge to make sure that those models are robustly aligned (Bowman et al., 2022). For example, we might hope to reliably know whether a model is being deceptive in order to achieve an instrumental goal. Importantly, deceptive alignment and robust alignment are behaviorally indistinguishable (Hubinger et al., 2024).

Aug 22, 2024 | lesswrong.com | Winnie Yang |Jojo Yang

As models grow increasingly sophisticated, they will surpass human expertise. It is a fundamentally difficult challenge to make sure that those models are robustly aligned (cite scalable oversight). For example, we might hope to reliably know whether a model is being deceptive in order to achieve an instrumental goal (Hubinger et al., 2024). Importantly, deceptive alignment and robust alignment are behaviorally indistinguishable (Hubinger et al., 2024).