arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2404.10552cs.CLcs.AI
收藏

通过上下文学习揭示基础大语言模型的滥用潜力

Unveiling the Misuse Potential of Base Large Language Models via In-Context Learning

Fudan University(复旦大学);University of California, Santa Barbara(加利福尼亚大学圣巴巴拉分校);Shanghai AI Laboratory(上海人工智能实验室)

Xiao Wang, Tianze Chen, Xianjun Yang, Qi Zhang, Xun Zhao, Dahua Lin

AI总结:

本研究揭示基础大语言模型可通过精心设计的上下文演示执行恶意指令,其风险水平与恶意微调模型相当,且极易被利用,呼吁重视基础大语言模型的安全防护。

摘要:

大语言模型(LLM)的开源加速了应用开发、创新与科学进步,这既包括在大规模数据集上预训练且未经过对齐的基础模型(base model),也包括经专门设计以符合伦理标准和人类价值观的对齐模型(aligned model)。普遍观点认为,基础大语言模型固有的指令遵循局限性可作为防范滥用的安全屏障,而我们的研究揭露了这一认知存在关键疏漏。通过部署精心设计的演示样例,本研究证明基础大语言模型能够有效解读并执行恶意指令。为系统评估这些风险,我们提出了一套全新的风险评估指标。实证结果显示,基础大语言模型的输出风险水平可与专为恶意目的微调的模型相当。这种漏洞既不需要专业知识也无需训练,几乎任何人都可利用,凸显了其重大风险以及立即关注基础大语言模型安全协议的紧迫性。

英文摘要:

The open-sourcing of large language models (LLMs) accelerates application development, innovation, and scientific progress. This includes both base models, which are pre-trained on extensive datasets without alignment, and aligned models, deliberately designed to align with ethical standards and human values. Contrary to the prevalent assumption that the inherent instruction-following limitations of base LLMs serve as a safeguard against misuse, our investigation exposes a critical oversight in this belief. By deploying carefully designed demonstrations, our research demonstrates that base LLMs could effectively interpret and execute malicious instructions. To systematically assess these risks, we introduce a novel set of risk evaluation metrics. Empirical results reveal that the outputs from base LLMs can exhibit risk levels on par with those of models fine-tuned for malicious purposes. This vulnerability, requiring neither specialized knowledge nor training, can be manipulated by almost anyone, highlighting the substantial risk and the critical need for immediate attention to the base LLMs' security protocols.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑