The Use of a Large Language Model for Cyberbullying Detection