Subtitle Milan, Italy, September 29–October 4, 2024, Proceedings, Part XXI. FALCON: Fair Active Learning for Content Moderation. - Generalizing Fairness to Generative Language Models via Reformulation of Non-discrimination Criteria.