SHIELD: Security against Harmful Injection Evaluation and Language Detection Leveraging Large Language Model To use this dataset, you must cite our paper "SHIELD: Security against Harmful Injection ...