Portrait of Ed

Edward Williams

PhD student in machine learning at the University of Oxford. I'm broadly interested in the statistical mechanics of learning and how tools like Gaussian processes and kernels can help us understand neural networks. More generally, I care about representation learning and meta-learning in high-dimensional settings.

You can reach me at [first name].[surname] [at] cs.ox.ac.uk.

Research

I think about how learning dynamics shape internal representations in overparameterised models, with a particular interest in:

  • Statistical mechanics of learning and generalisation
  • Representation learning for complex, high-dimensional data
  • Gaussian processes, kernels, and Bayesian viewpoints

Hobbies

Outside of machine learning, I’m interested in mathematics and theoretical physics more broadly. I also hugely enjoy cooking, photography, and hiking, and playing football, squash, and pool.