Learning Strategies Evolution
John discusses the importance of combining fine-tuning with in-context learning for model development. He emphasizes the need for models to introspect on their knowledge and actively seek new information. The conversation delves into the evolution of learning strategies towards more efficient and effective methods in long-horizon tasks.In this clip
From this podcast

Dwarkesh Podcast
John Schulman (OpenAI Cofounder) - Reasoning, RLHF, & Plan for 2027 AGI
Related Questions