The Alignment Problem from a Deep Learning Perspective

Authors

Richard Ngo, Lawrence Chan, Sören Mindermann

Abstract

This paper discusses the alignment problem in deep learning systems and proposes several approaches to address it.

Publication Details

Published:

March 15th, 2022

Venue:

arXiv

Added to AI Safety Papers:

March 16th, 2025

Metadata

Tags:

AlignmentDeep LearningAI Safety

Original Paper:

Link