Data prototypes for interpretability

Paper Code

Experiments with fixed features here.

Experiments with end-to-end learned model (fork of prototypical-networks) here.

Updated: