# Jailbreak Attack Detection

A security protection model fine-tuned based on DeBERTa-v3-large, used to detect unsafe content in conversations with large language models and prevent jailbreak attacks.