One paper on Safeguard for Large Language Model is accepted by ACL 2025