Sahildhonde-9/INJEXIS-Prompt-Injection-Dataset

The RedLockX Dataset is a large-scale curated security dataset designed for evaluating and training AI systems against adversarial threats such as prompt injection, jailbreak attempts, system prompt leakage, and LLM manipulation attacks. It contains structured real-world and synthetic attack patterns used in modern AI red-teaming. 📌 Dataset Overview ✔ 109,000+ labeled adversarial & safe samples ✔ Multi-category threat… See the full description on the dataset page: https://huggingface.co/datasets/Sahildhonde-9/INJEXIS-Prompt-Injection-Dataset.