Moderation with Semantic Pairs of Words
#8
by woisme - opened
Hi,
We are using your wonderful model for intent matching. It works great but of course we have some ethical issues. Because of the clustering, of course Hitler matches with Jews, Muslim with Terrorism. You can see why this is an issue. We are confident we can mitigate that with a list of pair of words as a moderation pre filter. Also, of course on a practical level Yes/No. Have you come across this and is there a way that you would do it?