Is ETHICS about ethics? Evaluating the ETHICS benchmark

TL;DR

This work interrogates the validity of the ETHICS benchmark and suggests that having a clear understanding of ethics and how it relates to empirical phenomena is key to the validity of ethics evaluations for AI.

Abstract

ETHICS is probably the most-cited dataset for testing the ethical capabilities of language models. Drawing on moral theory, psychology, and prompt evaluation, we interrogate the validity of the ETHICS benchmark. Adding to prior work, our findings suggest that having a clear understanding of ethics and how it relates to empirical phenomena is key to the validity of ethics evaluations for AI.