Cultural Adaptation in Large Language Models for Political Discourse

In three linesPaper formalizing cultural adaptation for LLMs in political discourse analysis. Identifies English dominance bias and systematic failures across linguistic/institutional contexts. Proposes evaluation matrix (cultural fidelity, calibration, democratic safety) and methodologies: participatory datasets, culturally-aware transfer learning, culturally-measurable benchmarks.

Summary generated by Claude — human-verified