Any thoughts on how granular citations for AI generated taxonomy/ontology objects should be? I'm seeing multiple options, and wondering if there is consensus. I.e. Would you prefer to cite the system the AI is used in, or the specific model and version?