license: cc-by-4.0
task_categories:
- text-classification
language: - en
tags: - jailbreaking
- injection
- LLM
- response
size_categories: - 10K<n<100K
This dataset is a part of our SHIELDING dataset that progressed to LLaMA-3 and checked each response humanly, as a part of our paper "Towards Secure Prompt Processing: A Unified Framework to Detect, Sanitize, and Prevent Adversarial Prompts via Dual-optimized Threshold-aware Learning".