AI.info

Brian Christian

Explore Brian Christian on AI.info.

• Reward Models Inherit Value Biases from Pretraining