It would be useful if it was possible to download just one csv file for all versions of a task (that are set up as separate nodes in experiment builder). At the moment it can be a bit cumbersome to identify which of many csv files are for different versions of the same task and them processing them into a single dataset.