The wildcard "attribute_eyebrows" supplements the face shape wildcards. It contains more than 400 attribute combinations. As my PROMPT test "(Ink illustration on parchment):1.5 of ((eye with (__attribute_eyebrows__))) {male|female}" shows, the differences are most obvious when drawings are created. In combination with many other prompts, less clear differences are generated. The different models and the sampling method also influence the representation.
It is therefore necessary to try out the weighting and the sampling method in the different models for each production.