Recent contributions are on Anthropic’s Frontier Red Team blog, focusing on building defensive cybersecurity capabilities, evaluating Claude’s performance in CTF competitions, and collaborating with Brian Singer to measure LLMs’ network attack capabilities. Covered by Fortune and Axios.
Research Experience
Currently part of Anthropic's Frontier Red Team, evaluating, improving, and discussing the cybersecurity capabilities of large language models. Previously served in several unique roles in the Air Force, separated as a Captain (O3).
Education
PhD from Carnegie Mellon University (CMU), advised by Lujo Bauer; MS in Information Security from CMU INI; BS in Computer Science from the US Air Force Academy.
Background
Research interests include cybersecurity capabilities of large language models, machine learning for malware detection, cooperative reinforcement learning, and robust machine learning. Gave a talk at DEF CON 33.