arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2501.09431cs.AIcs.CLcs.CRcs.CY
收藏

负责任的大型语言模型综述:内在风险、恶意使用与缓解策略

A Survey on Responsible LLMs: Inherent Risk, Malicious Use, and Mitigation Strategy

Tsinghua University(清华大学);Huazhong University of Science and Technology(华中科技大学)

Huandong Wang, Wenjie Fu, Yingzhou Tang, Zhilong Chen, Yuxi Huang, Jinghua Piao, Chen Gao, Fengli Xu, Tao Jiang, Yong Li

AI总结:

本文综述LLM的内在风险与恶意使用问题,按开发使用四阶段梳理隐私、幻觉、价值、毒性及越狱防御等缓解技术,并提出统一的负责任LLM框架。

摘要:

尽管大型语言模型(LLM)在支持众多实际应用和产生积极社会影响方面展现出巨大潜力,但它们仍面临隐私泄露、幻觉输出和价值错位等内在风险方面的严峻挑战,并且在被越狱后可能被恶意用于生成有毒内容和达成不道德目的。因此,在本综述中,我们全面回顾了旨在缓解这些问题的最新进展,并按照LLM开发与使用的四个阶段进行组织:数据收集与预训练、微调与对齐、提示与推理、后处理与审计。我们详细阐述了在隐私保护、幻觉减少、价值对齐、毒性消除和越狱防御方面提升LLM性能的最新进展。与以往聚焦负责任LLM单一维度的综述不同,本综述提出了一个涵盖这些不同维度的统一框架,为增强LLM以更好地服务实际应用提供了全面视角。

英文摘要:

While large language models (LLMs) present significant potential for supporting numerous real-world applications and delivering positive social impacts, they still face significant challenges in terms of the inherent risk of privacy leakage, hallucinated outputs, and value misalignment, and can be maliciously used for generating toxic content and unethical purposes after been jailbroken. Therefore, in this survey, we present a comprehensive review of recent advancements aimed at mitigating these issues, organized across the four phases of LLM development and usage: data collecting and pre-training, fine-tuning and alignment, prompting and reasoning, and post-processing and auditing. We elaborate on the recent advances for enhancing the performance of LLMs in terms of privacy protection, hallucination reduction, value alignment, toxicity elimination, and jailbreak defenses. In contrast to previous surveys that focus on a single dimension of responsible LLMs, this survey presents a unified framework that encompasses these diverse dimensions, providing a comprehensive view of enhancing LLMs to better serve real-world applications.

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。

↑