I am a programmer and researcher who wants to make computers more fun and empowering for students, researchers, and engineers. I thought the field was stuck, so I moved from computer science to neuroscience in search of ideas for making computers smarter. Then, to my surprise, the computers went and got smarter.

Current project

Can we make it more fun to explore the space of new AI models? Could exploring Deep Learning architectures and training regimes become more fun than playing Minecraft?

So far, two ML libraries have spun out from this project:

  • Vexpr: use Lisp-like code transformation to make models readable, fast, and visualizable
  • rows2prose: visualize models by rendering scalars into styled text


Expressions are Pragmatic Model Visualizations

What happens when you vectorize wide PyTorch expressions?

Gaussian Processes Extrapolate, Sometimes in Goofy Ways

Maybe Bayesian Optimization Should Be Harder, Not Easier

Imagine A Deep Network That Performs Successive Cheap Queries On Its Input

Bayesian Optimization Is More Basis-Dependent Than You Might Think

Likely ≠ Typical: A Viewpoint On Why We Perturb Neural Networks

Some “Causal Inference” intuition

See all posts

Select talks / presentations

Intro to Grid Cells + Quickly Forming Structured Memories

Testing a possible explanation for grid cell distortions

Journal Club: Hinton’s GLOM + Numenta’s TBT (after a year without a haircut)

Using grid cells as a prediction-enabling basis

The Minimum Description Length Principle, sparsity, and quantization


(Photo credit: Rosanne Liu, 2023)


Hippocampal Spatial Mapping As Fast Graph Learning
Marcus Lewis
Poster at 30th Annual Computational Neuroscience Meeting (2021)
Efficient and flexible representation of higher-dimensional cognitive variables with grid cells
Mirko Klukas, Marcus Lewis, Ila Fiete
PLOS Computational Biology (2020)
A Framework for Intelligence and Cortical Function Based on Grid Cells in the Neocortex
Jeff Hawkins, Marcus Lewis, Mirko Klukas, Scott Purdy, Subutai Ahmad
Front. Neural Circuits (2019)
Locations in the Neocortex: A Theory of Sensorimotor Object Recognition Using Cortical Grid Cells
Marcus Lewis, Scott Purdy, Subutai Ahmad, Jeff Hawkins
Front. Neural Circuits (2019)

Other projects, big and small

Using Grid Cells for Coordinate Transforms
Marcus Lewis
Poster, Grid Cell Meeting 2018, UCL, London, England
Grid cells: Visualizing the CAN model
A weekend in April 2017
See HTM run: Stacks of time series
Written while living in hostels. February 2016
A visual running environment for HTM
Collaboration with Felix Andrews. November 2015


Older posts | Google Scholar | LinkedIn | Mastodon