When AI Spills Its Secrets: Detecting System Prompt Leaks in Large Language Models

Published: 18 Nov 2025, Last Modified: 18 Nov 2025NeurIPS-25 EducationEveryoneRevisionsBibTeXCC BY 4.0

Keywords: LLM safety, system prompt, text similarity

Cover Page: pdf

Educational Material: zip

Submission Number: 33

Loading