Large language models (LLMs) are reported to be partial to certain cultures
owing to the training data dominance from the English corpora. Since
multilingual cultural data are often expensive to collect, existing efforts
handle this by prompt engineering or culture-specific pre-training. However,
they might overlook the knowledge deficiency of low-resource culture and
require extensive computing resources. In this paper, we propose CultureLLM, a
cost-effective solution to incorporate cultural differences into LLMs.
CultureLLM adopts World Value Survey (WVS) as seed data and generates
semantically equivalent training data via the proposed semantic data
augmentation. Using only 50 seed samples from WVS with augmented data, we
fine-tune culture-specific LLMs and one unified model (CultureLLM-One) for 9
cultures covering rich and low-resource languages. Extensive experiments on 60
culture-related datasets demonstrate that CultureLLM significantly outperforms
various counterparts such as GPT-3.5 (by 8.1%) and Gemini Pro (by 9.5%) with
comparable performance to GPT-4 or even better. Our human study shows that the
generated samples are semantically equivalent to the original samples,
providing an effective solution for LLMs augmentation.

提出了一种经济高效的解决方案 CultureLLM，利用 World Value Survey (WVS) 作为种子数据，通过提议的语义数据增强方法生成语义上等价的训练数据，然后使用这些数据对文化特定的 LLMS 进行微调，以及一个统一模型 (CultureLLM-One) 来覆盖 9 种文化。实验结果表明 CultureLLM 在各种文化相关数据集上的性能明显优于其他对比模型，例如 GPT-3.5（8.1%）和 Gemini Pro（9.5%），且与 GPT-4 相当甚至更好。人类研究结果显示生成的样本在语义上与原样本相等，为 LLMs 的增强提供了有效的解决方案。