Samuel J. Bell

Machine learning reproducibility, robustness and fairness.


I’m Sam đź‘‹, a final year machine learning PhD student in the [email protected] group at the University of Cambridge, supervised by Prof. Neil Lawrence.

My research focuses on understanding how contemporary machine learning models actually work. Broadly, I’m interested in reproducibility and metascientific research, tooling and theory for improving the efficiency of machine learning research, and understanding algorithmic biases and questions of fairness.

Previously, I was a placement student at The Alan Turing Institute, did my master’s in deep learning for natural language processing at the Cambridge Computer Laboratory, and did a bachelor’s in computer science at the University of Manchester.

In between, I’ve worked on algorithmic bias at FAIR; simulated financial crises in market risk at Goldman Sachs; built new retail banks at Thought Machine, and developed next generation credit scores at Credit Kudos.

I’m also the Founder and Chair of The Preptrack Foundation, a registered charity building technology for HIV prevention. Our first app, Preptrack, helps people who use PrEP, the medication that eliminates risk of HIV infection.

If you’re interested in what I do, research collaborations or opportunities, please do drop me a line. My email is sjb326 [at] cam [dot] ac [dot] uk. I’d love to hear from you.

Selected publications

  1. Modeling the Machine Learning Multiverse
    Samuel J. Bell, Onno Kampman, Jesse Dodge, and Neil Lawrence
    In Advances in Neural Information Processing Systems 2022
  2. The Effect of Task Ordering in Continual Learning
    Samuel J. Bell, and Neil Lawrence
  3. Behavioral Experiments for Understanding Catastrophic Forgetting
    Samuel J. Bell, and Neil Lawrence
    In AI Evaluation Beyond Metrics (EBeM) Workshop, IJCAI 2022
  4. ICLR
    Perspectives on Machine Learning from Psychology’s Reproducibility Crisis
    Samuel J. Bell, and Onno Kampman
    In Science and Engineering of Deep Learning (SEDL) Workshop, ICLR 2021