Tolga

This can also be used to "disable" some features (e.g. by setting a numerical feature to 0) to analyze how model scores change in the facets dive scatter plot.

Hi shlomihod, Thanks for reaching out. The code for UCI census web demo is [here](https://github.com/PAIR-code/what-if-tool/blob/master/wit_dashboard/demo/wit-age-demo.html), this and other files in the demo folder can be useful for your goal. For...

Interesting, as James mentioned, would it be possible for you to provide us a jupyter/colab notebook or a code snippet that produces this output?

in config.lua in function getConfig change the following lines to: ``` ... optimMethod = optim.sgd, optimConfig = {learningRate = 1e-3, momentum = 0.9}, ... ``` check https://github.com/torch/optim/blob/master/sgd.lua for more details.

Hi Breakend, For now, you can check the python tutorial notebook on how to capture racial bias direction and generate analogies.