Hi
this SVM module has two different modalities. If only train data is provided, it creates the model file (as you have found). If instead you run it with test data AND a model file, it will generate prediction results which you can use in the PredictionResultsViewer visualizer. If you give it train and test data (but no model) it will generate a model and then use that to create prediction results.
I believe the model file is an R binary format, and we don't have any way to visualize that in GenePattern, but you could probably load it into R to look at it but we have no documentation on what that might look like.
I am guessing that what you really want is prediction results, so you can either reuse that model file with another dataset, or split your dataset into train and test portions (e.g. using the SplitDatasetTrainTest module) to generate predictions on the test portion.
Hope this helps
Ted