---
id: 20260711-T0-10
title: "PLURAL 数据集发布：全球价值观对齐，覆盖 90 多个区域"
title_en: "PLURAL Dataset Released: Global Value Alignment Covering 90+ Regions"
url: https://ai.daily.yangsir.net/daily/20260711-T0-10
issue_date: 2026-07-11
publish_date: 2026-07-10T04:00:00.000Z
category: research
source_name: "arXiv cs.CL (NLP)"
source_url: https://arxiv.org/abs/2607.08034
---

# PLURAL 数据集发布：全球价值观对齐，覆盖 90 多个区域

针对大语言模型（LLM）过度反映西方价值观的问题，研究人员发布了 PLURAL 数据集。这是一个大规模、以价值观为核心的偏好数据集，基于整合价值观理论（Integrated Value Theory）构建，覆盖全球 90 多个国家和地区。该数据集可用于训练和评估 LLM 的价值观对齐能力，帮助模型更好地理解和尊重不同文化的价值体系。

## English Version

**PLURAL Dataset Released: Global Value Alignment Covering 90+ Regions**

Addressing the issue that LLMs disproportionately reflect Western values, researchers have released the PLURAL dataset. This large-scale, value-focused preference dataset is grounded in the Integrated Value Theory and covers over 90 countries and regions worldwide. It can be used to train and evaluate LLMs' value alignment capabilities, helping models better understand and respect diverse cultural value systems.

---

**来源**：[arXiv cs.CL (NLP)](https://arxiv.org/abs/2607.08034)

**详情页**：https://ai.daily.yangsir.net/daily/20260711-T0-10

---

*智语观潮 · Daily — https://ai.daily.yangsir.net/llms.txt*