Hate Personified: Investigating the role of LLMs in content moderation