AI Snips
Chapters
Transcript
Episode notes
Deep RL Generalization Challenges
- Generalization and robustness are major challenges in deep RL.
- Small input shifts can cause policies to fail, limiting real-world deployment.
Inspiration from PhD Exam Pressure
- Natasha Jaques struggled with a hard question from her PhD committee about social and emotional intrinsic motivations for agents.
- She spent nearly 24 hours crafting the idea of rewarding agents for having causal influence on others' actions.
Dig Into Model Behaviors
- Deeply analyze and visualize model behaviors beyond charts.
- Audit RL policies carefully to uncover unexpected emergent strategies.