I’m a Talos Fellow at SaferAI, developing better quantification tools for risk estimation of LLMs on cybersecurity. I have ten years of research experience (PhD, two postdoctoral positions and MPhil) and have worked on AI Safety (Technical & Policy).