Seth Nabarro

pic.jpg

I am a machine learning PhD student at Imperial College London, supervised by Mark van der Wilk and Andrew Davison.

I studied physics at Imperial for my undergrad degree, and did a master’s in computational statistics and machine learning at University College London. Before starting my PhD I was a research engineer at Graphcore, looking at hardware acceleration of probabilistic machine learning. More recently, I spent some time as a research intern at Google DeepMind.

Among other things, I am interested in local learning rules for deep learning, continual learning, multi-robot learning, and generative modelling.

news

Feb 01, 2026 Our paper A Distributed Gaussian Process Model for Multi-Robot Mapping was accepted at ICRA 2026!
Nov 04, 2025 Submitted my thesis, Towards Deep and Distributed Bayesian Learning!
Jan 05, 2025 Did an internship at Google DeepMind in Zürich August to December 2024, with Mark Collier, Shawn Wang and Efi Kokiopoulou.
Aug 01, 2024 Presented Learning in Deep Factor Graphs with Gaussian Belief Propagation at ICML in Vienna.
Nov 07, 2023 Presented some recent papers on forward gradients (Baydin et al., 2022, Ren et al., 2023) at reading group (slides).
Jun 30, 2023 Won Best Poster for Learning in deep factor graphs with Gaussian belief propagation at the Imperial College Computing Summer Conference, 2023.
May 26, 2023 Chosen as a Top Reviewer for UAI 2023.
Aug 05, 2022 Presented Data augmentation in Bayesian neural networks and the cold posterior effect at UAI in Eindhoven.