A theoretical resource for somewhat of a clean code. I keep a public version for most of the codes I wrote during my MSc at Gibhub. It is designed more as a reference for anyone interested in implementing a few things rather than as code that is designed to be run.


Recent Evolution of QA Datasets and Going Forward [Post]
Deep Neural Network Loss Surface Exploration [Post]
Metacademy: Dynamical Systems for Machine learning [Post]


Notes on Gaussian Processes & code [Note, Code]

Notes on Dirichlet Processes & CRP code [Note, Code]

Research codes

  • See my Github