Fernando Ruiz Mazo

Incoming PhD Student, University of Cambridge

fr440 [AT] cam.ac.uk

Bio

I am an incoming PhD student in Pure Mathematics and Mathematical Statistics at the University of Cambridge, supervised by Sergio Bacallado and advised by Richard Nickl. I am a QRT Labs Fellow, a Rafael del Pino Excellence Scholar, and a La Caixa Foundation Excellence Fellow.

My research lies at the intersection of pure mathematics, statistics, and machine learning. I am currently interested in understanding in-context learning in transformers -- in particular on when and under which conditions we can guarantee it will happen.

Currently, I am completing an MPhil in Mathematics at Cambridge under the supervision of Sergio Bacallado, doing research on similar topics. Previously, I was a research assistant at Aalto University under the supervision of Vikas Garg, working on topics related to generative models, uncertainty quantification, and prequential statistics. I received my BSc in Mathematics from the University of Seville, with a final dissertation on topological data analysis and quiver theory, written during my last year as an exchange student at the University of Warwick.

Publications

Frozen Priors, Fluid Forecasts: Prequential Uncertainty for Low-Data Deployment with Pretrained Generative Models

Fernando Ruiz Mazo, Vikas Garg

ICLR 2026