Super. Thanks, Marcos. It was indeed the descriptions that I was curious about.

Apart from the one that detects presence of 's' in a word, could you also provide a few other usecases for implementing custom mention_extractors?