Machine Learning Safety

Daniel Hendrycks

EECS Department, University of California, Berkeley

Technical Report No. UCB/EECS-2022-253

December 1, 2022

This publication is archived. It is kept only for reference purposes, so it is no longer being updated and may not meet accessibility standards. If you need this content in a different format, please email webteam@eecs.berkeley.edu.

http://www2.eecs.berkeley.edu/Pubs/TechRpts/2022/Archive/EECS-2022-253.pdf

Machine learning (ML) systems are rapidly increasing in size, are acquiring new capabilities, and are increasingly deployed in high-stakes settings. To address the growing need for safe ML systems, I first discuss works towards making systems perform reliably. Thereafter I discuss works towards making systems act in accordance with human values. In closing I discuss open problems in making ML systems safer.

BibTeX citation:

@phdthesis{Hendrycks:EECS-2022-253,
    Author= {Hendrycks, Daniel},
    Title= {Machine Learning Safety},
    School= {EECS Department, University of California, Berkeley},
    Year= {2022},
    Month= {Dec},
    Url= {http://www2.eecs.berkeley.edu/Pubs/TechRpts/2022/EECS-2022-253.html},
    Number= {UCB/EECS-2022-253},
    Abstract= {Machine learning (ML) systems are rapidly increasing in size, are acquiring new capabilities, and are increasingly deployed in high-stakes settings. To address the growing need for safe ML systems, I first discuss works towards making systems perform reliably. Thereafter I discuss works towards making systems act in accordance with human values. In closing I discuss open problems in making ML systems safer.},
}

EndNote citation:

%0 Thesis
%A Hendrycks, Daniel 
%T Machine Learning Safety
%I EECS Department, University of California, Berkeley
%D 2022
%8 December 1
%@ UCB/EECS-2022-253
%U http://www2.eecs.berkeley.edu/Pubs/TechRpts/2022/EECS-2022-253.html
%F Hendrycks:EECS-2022-253