Freddie Bickford Smith

Email | GitHub | Google Scholar | Twitter

I'm a PhD student in the RainML lab (OxCSML) at Oxford, supervised by Tom Rainforth and Adam Foster. My research focuses on intelligent data acquisition for machine learning [7-9]. I have also worked on an approximate-inference scheme to allow neural networks to learn more effectively across sequences of tasks [6]. Previously, during my MSc at UCL, I collaborated with Brad Love, Brett Roads and Ed Grefenstette on a project aimed at understanding neural networks from a cognitive-science perspective [3-5]. As an undergraduate I studied mechanical engineering at Bristol, working on carbon-fibre composites [2] with Fabrizio Scarpa and on 3D printing [1] with Ben Hicks.

Research

  1. Making better use of unlabelled data in Bayesian active learning
    F Bickford Smith, A Foster, T Rainforth
    International Conference on Artificial Intelligence and Statistics (AISTATS), 2024
  2. Modern Bayesian experimental design
    T Rainforth, A Foster, D Ivanova, F Bickford Smith
    Statistical Science, 2024
  3. Prediction-oriented Bayesian active learning
    F Bickford Smith*, A Kirsch*, S Farquhar, Y Gal, A Foster, T Rainforth
    International Conference on Artificial Intelligence and Statistics (AISTATS), 2023
  4. Continual learning via sequential function-space variational inference
    T Rudner, F Bickford Smith, Q Feng, YW Teh, Y Gal
    International Conference on Machine Learning (ICML), 2022
  5. Understanding top-down attention using task-oriented ablation design
    F Bickford Smith, B Roads, X Luo, B Love
    arXiv, 2021
  6. The perceptual boost of visual attention is task-dependent in naturalistic settings
    F Bickford Smith, X Luo, B Roads, B Love
    "Bridging AI and Cognitive Science", ICLR, 2020
  7. When is visual attention useful?
    F Bickford Smith
    MSc thesis, University College London, 2019
  8. An investigation into methods of joining fibre-reinforced plastics for Formula Student
    F Bickford Smith, C Lo, A Vossos
    MEng thesis, University of Bristol, 2017
  9. Applications of 3D printing for Formula Student: optimisation of the impact attenuator
    F Bickford Smith
    BEng thesis, University of Bristol, 2016