Hi, I'm Leo.
I'm an AI safety researcher and founder of a new non-profit that's building tools to accelerate AI safety research. The long-term goal is to automate AI safety research to ensure it can co-scale with the rapid advancement of AI capabilities. The short-term goal is to build dev tools that help today's researchers move faster.
Before transitioning to AI safety with scheming evals and interp, I worked as an engineer on drones and orbital rockets, then as an ML researcher in computational neuroscience.
Based in San Francisco. Reach out if you want to help close the gap between capabilities and safety.