AI.info

Brian Hu

Explore Brian Hu on AI.info.

• Aligning Machiavellian Agents: Behavior Steering via Test-Time Policy Shaping