This repository contains the file notebook_fig4.ipynb, which is a jupyter notebook that accompanies Figure 4 of the paper: A Review of the Gumbel-max Trick and its Extensions for Discrete Stochasticity in Machine Learning.
This notebook can be used to gain insights in the relations between Gumbel-max and Gumbel-softmax samples, generated from unnormalized logits or normalized probabilities. It has been tested on a Windows system in an anaconda environment with Python 3.6.7, numpy 1.17.3 and scipy 1.5.4.
Along with this notebook, you might find the following interactive demo useful as well: https://iamhuijben.github.io/gumbel_softmax_sampling.html
Please cite the following paper if you find these demos useful in your own work.
@article{
huijben2022,
author={Iris A.M. Huijben and Wouter Kool and Max B. Paulus and Ruud J.G. van Sloun},
journal={IEEE Transactions on Pattern Analysis and Machine Intelligence},
title={A Review of the Gumbel-max Trick and its Extensions for Discrete Stochasticity in Machine Learning},
year={2022},
doi={10.1109/TPAMI.2022.3157042},
}