大模型的价值观受训练数据文化背景影响，东西方模型差异显著。
The Cultural Gene of Large Language Models: A Study on the Impact of Cross-Corpus Training on Model Values and Biases
- 构建文化探针数据集，对比中西方模型在个体主义与权力距离上的倾向
- GPT-4偏向个人主义低权力距离，ERNIE Bot则倾向集体主义高权力距离
- 模型价值观与中美文化指标高度一致，提示需警惕算法文化霸权
大型语言模型（LLMs）在全球部署，但其潜在的文化与伦理假设仍缺乏深入研究。本文提出“文化基因”概念——指模型从训练语料中继承的系统性价值取向，并构建了包含200个提示的文化探针数据集（CPD），聚焦个体主义-集体主义（IDV）和权力距离（PDI）两个经典跨文化维度。通过标准化零样本提示，比较西方主导的GPT-4与东方主导的ERNIE Bot。人工标注显示两者在两维度上存在显著且一致的差异：GPT-4呈现个体主义与低权力距离特征（IDV ≈ 1.21；PDI ≈ -1.05），ERNIE Bot则表现为集体主义与高权力距离（IDV ≈ -0.89；PDI ≈ 0.76）；差异具有统计显著性（p < 0.001）。进一步计算文化契合度指数（CAI）发现，GPT-4更贴近美国文化（如IDV CAI ≈ 0.91；PDI CAI ≈ 0.88），ERNIE Bot更贴近中国文化（IDV CAI ≈ 0.85；PDI CAI ≈ 0.81）。对困境解决与权威判断的定性分析揭示了这些倾向在推理中的具体体现。结果表明，大模型是其文化语料的统计镜像，呼吁开展更具文化意识的评估与部署，以避免算法层面的文化霸权。
原文摘要 · Abstract (English)
Large language models (LLMs) are deployed globally, yet their underlying cultural and ethical assumptions remain underexplored. We propose the notion of a "cultural gene" -- a systematic value orientation that LLMs inherit from their training corpora -- and introduce a Cultural Probe Dataset (CPD) of 200 prompts targeting two classic cross-cultural dimensions: Individualism-Collectivism (IDV) and Power Distance (PDI). Using standardized zero-shot prompts, we compare a Western-centric model (GPT-4) and an Eastern-centric model (ERNIE Bot). Human annotation shows significant and consistent divergence across both dimensions. GPT-4 exhibits individualistic and low-power-distance tendencies (IDV score approx 1.21; PDI score approx -1.05), while ERNIE Bot shows collectivistic and higher-power-distance tendencies (IDV approx -0.89; PDI approx 0.76); differences are statistically significant (p < 0.001). We further compute a Cultural Alignment Index (CAI) against Hofstede's national scores and find GPT-4 aligns more closely with the USA (e.g., IDV CAI approx 0.91; PDI CAI approx 0.88) whereas ERNIE Bot aligns more closely with China (IDV CAI approx 0.85; PDI CAI approx 0.81). Qualitative analyses of dilemma resolution and authority-related judgments illustrate how these orientations surface in reasoning. Our results support the view that LLMs function as statistical mirrors of their cultural corpora and motivate culturally aware evaluation and deployment to avoid algorithmic cultural hegemony.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。