bengali

This paper introduces CulturalNB, a dataset of Bengali cultural question-answer pairs, and evaluates nine LLMs for cross-lingual cultural bias. Findings show that English prompting increases global narrative substitution and reduces local perspectives, revealing that cultural failures in LLMs are grounding and prioritization issues, not just missing knowledge.

This paper introduces MultiSoc-4D, a benchmark for diagnosing instruction-induced label collapse in LLMs annotating Bengali social media. It reveals that LLMs systematically prefer fallback labels, leading to under-detection of minority categories like hate speech and sarcasm.