For context, the author Chris Olah is a researcher at Open AI, who I think is one of the most clear and thoughtful writers in the space. He has a lot of articles on understanding ML models, and pioneers new methods of visualizations and interactivity to improve understandability of these models. If you're interested in the field I would highly recommend reading some of his articles on his blog or on distill.pub, a publication that he helped found with the express purpose of improving clarity of ML research.