The result of re-aplying a recognition model after having archived its corresponding transcription is invisible.

if you delete a transcription layer of an automatic transcription and reuse the same model to create a new one there is no chance to visualize it. the transcription layer will always remain archived.