Chasing Away Repetitive LLM Responses with Verbalized Sampling - Linear Digressions | Wave AI Podcast Notes