Notes

Rethinking deep vision interpretability

Explainability with attribution is limited to indicate the where. Concept extraction allows to capture the what.

Atomic auto-encoders for learning sparse representations

Using a specific type of auto-encoders to learn a sparse representation of your data

Job token permissions to access registries

How to allow a gitlab repository to access the container registry of another, private, repository

Regularized gradient descent

Plug-and-play algorithms

Changing surrounding delimiters in VIM

Python and Paraview for batch processing

One liners

One liner beauty for aggregating files and computing statistics