PRL:
We show Americans anonymized results from a variety of Large Language Models. We then ask them to assess the partisan slant of the models on 30 of the most salient issues in American politics in a head-to-head comparison.
e = get, head
Dive into said