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. Work cross-functionally with internal inference, product API, systems, and security teams, among others, and with CSP partners to stand up the full serving stack on new cloud platforms, resolve operational issues, and influence provider roadmaps.