Think Twice, Generate Once: Safeguarding by Progressive Self-Reflection

Hoang Phan, Victor Li, Qi Lei

Large language models (LLMs) have revolutionized natural language processing with their ability to generate coherent and contextually relevant text. However, their deployment raise…