We have updated our software for improved interpretability since this post was written. For the latest on this topic, read our new series on model-agnostic interpretability.  Assessing a model`s accuracy usually is not enough for a data scientist who wants to know more about how a model is working. Often