Machine learning (ML) systems are rapidly increasing in size, are acquiring new capabilities, and are increasingly deployed in high-stakes settings. As with other powerful technologies, safety for ML should be a leading research priority. In response to emerging safety challenges in ML, such as those introduced by recent large-scale models, we provide a new roadmap for ML Safety and refine the technical problems that the field needs to address. We present four problems ready for research, namely withstanding hazards ("Robustness"), identifying hazards ("Monitoring"), reducing inherent model hazards ("Alignment"), and reducing systemic hazards ("Systemic Safety"). Throughout, we clarify each problem's motivation and provide concrete research directions.
翻译:机器学习(ML)系统的规模正在迅速扩大,正在获得新的能力,并越来越多地部署在高占用环境中。与其他强大的技术一样,ML的安全应该成为主要的研究重点。为了应对ML中新出现的安全挑战,例如最近大规模模型引入的安全挑战,我们为ML安全提供了一个新的路线图,并完善了该领域需要解决的技术问题。我们提出了四个可供研究的问题,即常有的危害("交战")、识别危害("监测")、减少内在的模型危害("信号")和减少系统性危害(“系统安全” )。我们始终澄清每个问题的动机并提供具体的研究方向。