Aligning Large Language Models with Diverse Political Viewpoints
This addresses the issue of normative political stances in AI responses for users seeking unbiased political information, though it is incremental as it focuses on a specific domain and dataset.
The paper tackles the problem of political bias in large language models by aligning them with diverse political viewpoints from Swiss parliamentary candidates, resulting in models that generate more accurate political viewpoints compared to commercial models like ChatGPT.
Large language models such as ChatGPT exhibit striking political biases. If users query them about political information, they often take a normative stance. To overcome this, we align LLMs with diverse political viewpoints from 100,000 comments written by candidates running for national parliament in Switzerland. Models aligned with this data can generate more accurate political viewpoints from Swiss parties, compared to commercial models such as ChatGPT. We also propose a procedure to generate balanced overviews summarizing multiple viewpoints using such models. The replication package contains all code and data.