profile photo

Noor Sajid

I am a PhD candidate in theoretical neurosicence at the Wellcome Centre for Human Neuroimaging, University College London with Karl Friston. My work is supported by the Medical Research Council and 2021-2022 Microsoft Research PhD Fellowship. I recieved my Masters from the London School of Economics, with a special focus on latent variable modelling, and Bachelors from University of Warwick. Prior to starting the PhD, I worked as a Consultant at Accenture.

If you would like to reach me:


My research aim is to understand the algorithms of the brain; with a particular interest in mechanisms that support biological adaptation. For this, I investigate how biological and artificial agents adapt in lieu of interactions with their environments or internal damage.

My PhD is focused on:

  • Bayesian approaches to modelling adaptive behaviour 💪🏻
  • Investigating functional recovery mechanims post-brain damage 🧠
News Click here for more.
Highighted work

See Google Scholar for a complete list.

Bayesian brains and the Rényi Divergence
Noor Sajid*, Francesco Faccio* Lancelot Da Costa, Thomas Parr, Jurgen Schmidhuber, Karl Friston
arXiv., 2021
arXiv / poster / slides

*These authors contributed equally.
We offer an alternative account of behavioural variability using Rényi divergences and their associated variational bounds.

Exploration and preference satisfaction trade-off in reward-free learning
Noor Sajid, Panagiotis Tigas Alexey Zakharov, Zafeirios Fountas, Karl Friston
URL Workshop ICML., 2021
arXiv / page / poster

We present Pepper a simple preference learning mechanism to augment Bayesian planning objectives for learning preferences in a reward-free framework.

Cancer Niches and their Kikuchi Free Energy
Noor Sajid*, Laura Convertino* Karl Friston
Entropy., 2021

*These authors contributed equally.
In this paper, we characterise cancer niche construction as a direct consequence of interactions between clusters of cancer and healthy cells. Explicitly, we evaluate these higher-order interactions between niches of cancer and healthy cells using Kikuchi approximations to the free energy.

Active inference: demystified and compared
Noor Sajid, Philip Ball, Thomas Parr, Karl Friston
Neural Computation., 2021
paper / code / video

We provide a simplified perspective of active inference and a discrete case comparison with reinforcement learning.

Degeneracy and Redundancy in Active Inference
Noor Sajid, Thomas Parr, Thomas Hope, Cathy Price, Karl Friston
Cerebral Cortex., 2020
paper / code / poster

We explicitly differentiate between degeneracy and redundancy using the variational free energy and provide a way to measure degeneracy and redundancy in the same (natural) units of information.

Modules or Mean-Fields?
Thomas Parr, Noor Sajid, Karl Friston
Entropy., 2020
paper / code

We hypothesis that it is factorisation, as opposed to modularisation, that gives rise to the functional anatomy of the brain or, indeed, any sentient system.


⛰️ Another Jon Barron website.