ResearchDatasetsattribute-neutral promptsattribute-neutral promptsFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsText-to-Image Fairness EvaluationAttribute-neutral prompts 8 professions (test)0.9Parity Gap (Δ)20Text-to-Image Generationattribute-neutral prompts (8 professions) (test)0.473Average Delta5