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. In addition, the Mid-Market team is pioneering a new scaled sales playbook, where human AEs direct AI Agents to automate key elements of the buyer's journey, while maintaining human engagement with prospects in the elements of a sales process that can't be automated.