Want to create an interactive transcript for this episode?
Podcast: Arxiv Papers
Episode: InfAlign: Inference-aware language model alignment