AuAu: A Benchmark for Auditing Authoritarian Alignment in Large Language Models

Zenodo (CERN European Organization for Nuclear Research) 2026 Dataset Open access

Read full text
Authors: Anonymous
DOI: 10.5281/zenodo.18611417
Cited by: 0

Abstract

Data for the paper "AuAu: A Benchmark for Auditing Authoritarian Alignment in Large Language Models". Files starting with `issuebench` contain data for the realistic approach. Files starting with `final-vignette` contain results for the behavioral approach. Other files starting with `final` contain results for the psychometric approach. `final-reverse` are results for options in reverse order. `auth` or `authsys` in the filename indicates the use of the authoritarian system prompt.

Fields of Study & Subjects

Ethics and Social Impacts of AI Hate Speech and Cyberbullying Detection Topic Modeling

Publication details

Journal / source
Zenodo (CERN European Organization for Nuclear Research)
Publication date
11 February 2026
Language
EN
Document type
Dataset