Beyond prompt brittleness: Evaluating the reliability and consistency of political worldviews in LLMs
This addresses the need to assess whether LLMs embed stable political biases, which is crucial for users and developers concerned about model reliability in political contexts, though it is incremental as it builds on prior findings of left-liberal leanings.
The study tackled the problem of understanding the reliability and consistency of political worldviews in large language models (LLMs) by testing them on political statements from EU voting-advice questionnaires, finding that reliability increases with parameter count and larger models show mixed left-leaning and right-wing stances across different policy areas.
Due to the widespread use of large language models (LLMs), we need to understand whether they embed a specific "worldview" and what these views reflect. Recent studies report that, prompted with political questionnaires, LLMs show left-liberal leanings (Feng et al., 2023; Motoki et al., 2024). However, it is as yet unclear whether these leanings are reliable (robust to prompt variations) and whether the leaning is consistent across policies and political leaning. We propose a series of tests which assess the reliability and consistency of LLMs' stances on political statements based on a dataset of voting-advice questionnaires collected from seven EU countries and annotated for policy issues. We study LLMs ranging in size from 7B to 70B parameters and find that their reliability increases with parameter count. Larger models show overall stronger alignment with left-leaning parties but differ among policy programs: They show a (left-wing) positive stance towards environment protection, social welfare state and liberal society but also (right-wing) law and order, with no consistent preferences in the areas of foreign policy and migration.