Dataset and Lessons Learned from the 2024 SaTML LLM Capture-the-Flag Competition

arXiv:2406.07954v1 Announce Type: new
Abstract: Large language model systems face important security risks from maliciously crafted messages that aim to overwrite the system's original instructions or leak private data. To study this problem, we organized a capture-the-flag competition at IEEE SaTML 2024, where the flag is a secret string in the LLM system prompt. The competition was organized in two phases. In the first phase, teams developed defenses to prevent the model from leaking the secret. During the second phase, …

aim arxiv capture capture-the-flag competition cs.ai cs.cr data dataset flag ieee important instructions language large large language model leak lessons learned llm messages private private data problem risks secret security security risks study system systems