Excellent moss.

I’m Umais. This is my personal website, used mostly for housing short infrequent essays.

In real life, my most recent role has been as a research scientist in the Neuromorphic Computing Group at Huawei R&D, where I worked on developing generative modelling and AI algorithms inspired by theoretical neuroscience.

Prior to that, I completed a masters in Computational Statistics and Machine Learning (CSML), as well as a Bachelor’s in Physics, both from UCL. I have also previously worked as a data scientist in a credit management firm, a data analyst, an english tutor and a pizza boy - in no order of preference!

Papers as primary author

  • Sample as You Infer: Predictive Coding With Langevin Dynamics
    Umais Zahid, Qinghai Guo, Zafeirios Fountas; arXiv:2311.13664
  • Predictive Coding as a Neuromorphic Alternative to Backpropagation: A Critical Evaluation
    Umais Zahid, Qinghai Guo, Zafeirios Fountas; Neural Computation 2023
  • Curvature-Sensitive Predictive Coding with Approximate Laplace Monte Carlo
    Umais Zahid, Qinghai Guo, Zafeirios Fountas; arXiv:2303.04976

Books worth reading

  • Surfing Uncertainty by Andy Clark
    This was my first proper introduction to theories around the bayesian brain. (minus a handful of his great review papers). Clark doesn’t bother covering any of the mathematical formalities, which depending on your exposure to the theory may actually be a good thing.
  • What Is Real? The Unfinished Quest for the Meaning of Quantum Physics by Adam Becker
    This is an excellent example of how dominant personalities can commandeer a scientific field, and often at their detriment.
  • Behave - Robert Sapolsky
    This is about as encyclopedic as a popular science book can be without becoming a textbook, it touches upon work from psychology, sociology and neurobiology. Worth reading just so that you can be aware of how much of our behaviour is influenced by factors outside the realm of our ‘control’.
  • One Hundred Years of Solitude - Gabriel Garcia Marquez
  • Siddhartha - Herman Hesse