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. Make sure finance apps and agents have the platform support they need: SDKs and frameworks, scoped access, SOX guardrails, observability, and the capabilities finance needs that claude products do not provide out of the box.