In ParaView, each view that a user creates may display datasets in the pipeline via a representation. Formally, a representation is a vtkAlgorithm subclass that adapts data for rendering based on the type of view: a spreadsheet view uses a different algorithm to prepare data for display than a 3-D render-view. Besides adaptations for different view types, representations may adapt different types of input datasets. SMTK provides a representation, vtkSMTKResourceRepresentation, that adapts models, meshes, and other resources (any that provide tessllation data on their components) for display in 3-D render views.