Adam Shai comments on Graphical tensor notation for interpretability