arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2307.16680cs.LGcs.AIcs.CLcs.CRcs.CV
收藏

最先进生成模型可信度全景:综述与展望

On the Trustworthiness Landscape of State-of-the-art Generative Models: A Survey and Outlook

East China Normal University(华东师范大学);Alibaba Group(阿里巴巴集团);Xidian University(西安电子科技大学)

Mingyuan Fan, Chengyu Wang, Cen Chen, Yang Liu, Jun Huang

AI总结:

本文从隐私、安全、公平性和责任性四个维度系统综述了扩散模型与大语言模型的可信度威胁,构建可信度全景图并提出安全应用建议与未来研究方向。

摘要:

扩散模型和大语言模型已成为领先的生成模型,正在革新人类生活的诸多方面。然而,这些模型的实际部署也暴露了固有风险,使其负面侧面凸显出来,并引发了对其可信度的担忧。尽管关于这一主题的文献十分丰富,但专门深入探讨大规模生成模型与其可信度之间交叉领域的综合性综述仍然基本缺失。为填补这一空白,本文从四个基本维度考察了与这些模型相关的长期存在和新出现的威胁:1)隐私,2)安全,3)公平性,以及4)责任性。基于调查结果,我们构建了一张详尽的地图,勾勒出大型生成模型的可信度状况。此后,我们为未来配备大型生成模型的安全应用提供了实用建议和潜在研究方向,最终提升模型的可信度并使整个社会受益。

英文摘要:

Diffusion models and large language models have emerged as leading-edge generative models, revolutionizing various aspects of human life. However, the practical implementations of these models have also exposed inherent risks, bringing to the forefront their evil sides and sparking concerns regarding their trustworthiness. Despite the wealth of literature on this subject, a comprehensive survey specifically delving into the intersection of large-scale generative models and their trustworthiness remains largely absent. To bridge this gap, this paper investigates both the long-standing and emerging threats associated with these models across four fundamental dimensions: 1) privacy, 2) security, 3) fairness, and 4) responsibility. Based on the investigation results, we develop an extensive map outlining the trustworthiness of large generative models. After that, we provide practical recommendations and potential research directions for future secure applications equipped with large generative models, ultimately promoting the trustworthiness of the models and benefiting the society as a whole.

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。

↑