Metraipolis › archive › dc_0779bbe22ad8
Navigating the digital spectrum: Assessing political bias, stability, and downstream fairness in Large Language Models — arXiv:2609.08637
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.08637
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-16T12:49:09.364Z
- SHA-256 of the archived text
fdec95b18b1a3c38a877cac50a83b47d89a84918a74931435d9ff16a903104cd- Stable URL of this page
https://metraipolis.com/documento/dc_0779bbe22ad8
How to cite
Navigating the digital spectrum: Assessing political bias, stability, and downstream fairness in Large Language Models — arXiv:2609.08637 — arXiv (en), https://arxiv.org/abs/2609.08637 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_0779bbe22ad8 (Metraipolis archive dc_0779bbe22ad8, sha256 fdec95b18b1a3c38a877cac50a83b47d89a84918a74931435d9ff16a903104cd)
BibTeX
@misc{dc0779bbe22ad8,
title = {Navigating the digital spectrum: Assessing political bias, stability, and downstream fairness in Large Language Models — arXiv:2609.08637},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.08637},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_0779bbe22ad8; sha256 fdec95b18b1a3c38a877cac50a83b47d89a84918a74931435d9ff16a903104cd},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.08637. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Large Language Models are increasingly deployed as information intermediaries, yet measuring their political behavior remains fragile because questionnaire results mix model dispositions with measurement artifacts and response-elicitation biases. We introduce a robust Political Compass Test evaluation framework that samples 300 configurations across an eight-dimensional perturbation space varying language, framing, instructions, answer format, option order, and persona wording. We evaluate eight Gemma 3 and Qwen 3 models across 14 languages and three quantization levels, obtaining design-averaged political coordinates with quantified uncertainty. Most models lean Libertarian-Left on average, but instruction phrasing, language, and answer format significantly affect recovered coordinates. Cross-lingual differences primarily reflect coordinate drift rather than distinct cultural reasoning. Reverse-engineering the test also exposes axis-weighting imbalances and the collapse of degenerate responses toward the center, so near-origin estimates for the smallest models can reflect weak signal rather than centrism. Free-text reasoning and chat-then-classify elicitation alter recovered coordinates, and larger models show clearer persona separation, with a specific failure of the Authoritarian-Left persona to move most models in the intended social direction. In downstream tasks, persona effects are modest relative to model size and target group for hate-speech detection, while base and centrist prompts give the highest agreement for topic-level sentiment. Political role prompting therefore has measurable but task- and dataset-specific downstream effects. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-09 in cs.CL). Categorie arXiv: Computation and Language (cs.CL) ; Computers and Society (cs.CY). Pagina del lavoro: https://arxiv.org/abs/2609.08637 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-16T12:30:25.196Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-09?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).