Matrix training is (in my opinion) one of the coolest programming advancement our field has made in a long time. The precise organization of multiple exemplar targets to promote generalization is fascinating, and Clements and colleagues really show us how it is done in this article!