This research continues many of the directions our team worked on prior to Anthropic, including: GPT-3, Circuit-Based Interpretability, Multimodal Neurons, Scaling Laws, AI & Compute, Concrete Problems in AI Safety, and Learning from Human Preferences. Search is core to how Claude answers questions about the world, and the team owns both the problem and the solution end to end: growing the index, improving retrieval quality and running the serving stack at production scale.