Bucked Unicorn

I research training and evaluation methods for cybersecurity-focused AI models, with an emphasis on adversary emulation, agentic learning, and the reliable acquisition of operational security capabilities.