Summary
Felix Biggs is an applied scientist with nine years of experience translating foundational ML theory into practical, safety-minded AI systems. He completed a PhD at UCL that produced six first-author papers (including a NeurIPS Spotlight) proving novel, non-vacuous generalisation guarantees and proposing algorithms that reduce the need for conventional test-set validation. Felix has extended this foundation-first approach into mechanistic interpretability work and applied research at Secondmind and Wayve, focusing on active learning, time-series modelling, feasible set estimation, and diffusion models for regression. He’s pioneered methods that combine diffusion processes with LLM knowledge and developed latent-space core-set selection techniques for expensive engineering simulations. Comfortable moving between theory and engineering, he also has experience in teaching and community-driven projects (Recurse Center) that reflect a broad, systems-level curiosity. Based in London, he blends rigorous academic insight with product-focused research to make AI systems more predictable and deployable.
9 years of coding experience
1 year of employment as a software developer
BlueDot Impact Governance Course, BlueDot Impact Governance Course at AI Safety Fundamentals
B.A. (Hons) M.A. Master of Science - MSci Physics, B.A. (Hons) M.A. Master of Science - MSci Physics at University of Cambridge
University College London
English, Spanish, Chinese