AI Alignment: A Comprehensive Survey

التفاصيل البيبلوغرافية
العنوان: AI Alignment: A Comprehensive Survey
المؤلفون: Ji, Jiaming, Qiu, Tianyi, Chen, Boyuan, Zhang, Borong, Lou, Hantao, Wang, Kaile, Duan, Yawen, He, Zhonghao, Zhou, Jiayi, Zhang, Zhaowei, Zeng, Fanzhi, Ng, Kwan Yee, Dai, Juntao, Pan, Xuehai, O'Gara, Aidan, Lei, Yingshan, Xu, Hua, Tse, Brian, Fu, Jie, McAleer, Stephen, Yang, Yaodong, Wang, Yizhou, Zhu, Song-Chun, Guo, Yike, Gao, Wen
سنة النشر: 2023
المجموعة: Computer Science
مصطلحات موضوعية: Computer Science - Artificial Intelligence
الوصف: AI alignment aims to make AI systems behave in line with human intentions and values. As AI systems grow more capable, so do risks from misalignment. To provide a comprehensive and up-to-date overview of the alignment field, in this survey, we delve into the core concepts, methodology, and practice of alignment. First, we identify four principles as the key objectives of AI alignment: Robustness, Interpretability, Controllability, and Ethicality (RICE). Guided by these four principles, we outline the landscape of current alignment research and decompose them into two key components: forward alignment and backward alignment. The former aims to make AI systems aligned via alignment training, while the latter aims to gain evidence about the systems' alignment and govern them appropriately to avoid exacerbating misalignment risks. On forward alignment, we discuss techniques for learning from feedback and learning under distribution shift. On backward alignment, we discuss assurance techniques and governance practices. We also release and continually update the website (www.alignmentsurvey.com) which features tutorials, collections of papers, blog posts, and other resources.
Comment: Continually updated, including weak-to-strong generalization and socio-technical thinking. 58 pages (excluding bibliography), 801 references
نوع الوثيقة: Working Paper
URL الوصول: http://arxiv.org/abs/2310.19852
رقم الأكسشن: edsarx.2310.19852
قاعدة البيانات: arXiv