Hands-on expertise in one or more of the following domains: offensive or defensive cyber operations (including vulnerability research, reverse engineering, and ICS/OT security), cryptography and cryptanalysis, intelligence analysis and ISR (all-source, SIGINT, GEOINT), mission and weapons systems (command and control, space, electronic warfare, autonomy), CBRN and nonproliferation, or accreditation of classified systems (RMF, ATO, cross-domain solutions). 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.