Each file has a number corresponding to the identifier of the job whose results it contains. These are CSV formatted.