Cross-Lingual Consistency in Language Models

This field asks whether language models preserve answers, reasoning, and safety behavior when the same meaning is expressed in different languages, scripts, or dialects. Researchers use parallel question-answering, factual-recall, reasoning, and safety benchmarks, measuring both accuracy and agreement across language versions. Methods include multilingual fine-tuning and reinforcement learning, representation alignment, preference optimization, and inference-time steering. Results show persistent gaps, especially for lower-resource languages, while leaving an important tension: models should be consistent on shared facts and safety rules without erasing culturally appropriate differences.