PLURAL: A Global Dataset for Value Alignment
Quick Answer
PLURAL is a new dataset aimed at improving value alignment in large language models, featuring ~500,000 preference triplets from 20 countries.
Quick Take
PLURAL is a new dataset aimed at improving value alignment in large language models, featuring ~500,000 preference triplets from 20 countries. It enhances cultural representation, reducing alignment error by up to 27.7% compared to existing baselines and is validated through automated and human evaluations.
Key Points
- PLURAL is based on the Integrated Values Survey, covering 92 countries.
- Dataset includes ~500,000 preference triplets reflecting diverse cultural values.
- Training on PLURAL improves alignment with cultural profiles by up to 27.7%.
- Evaluated by 176 human judges in India, Brazil, and Japan for representativeness.
- Provides a scalable resource for pluralistic alignment in AI systems.
Paper Resources
📖 Reader Mode
~2 min readAbstract:Large language models (LLMs) are used worldwide, yet disproportionately reflect Western values, limiting their ability to represent diverse value systems. We introduce PLURAL, a large-scale, value-focused preference dataset grounded in the Integrated Values Survey (IVS), a nationally representative survey spanning 92 countries. Using a two-stage generation pipeline, we transform survey responses into synthetic preference triplets that preserve normative value signals while producing realistic scenarios. We release an initial version of PLURAL containing ~500,000 preference triplets representing people in 20 diverse countries. We evaluate PLURAL in three ways: (i) dataset-level validation showing that it preserves both cross-country value differences and within-country diversity from the original survey; (ii) automated evaluation showing that training on PLURAL improves alignment with target countries' cultural profiles, reducing mean absolute error by up to 27.7% relative to strong baselines; and (iii) blind human evaluation with 176 evaluators in India, Brazil, and Japan, who judge PLURAL-aligned responses as more representative of their national values. Together, these results show that PLURAL contains learnable signal for value steering, offering a scalable resource for pluralistic alignment. Dataset: this https URL
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Computers and Society (cs.CY) |
| Cite as: | arXiv:2607.08034 [cs.CL] |
| (or arXiv:2607.08034v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2607.08034 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Dhruv Agarwal [view email]
[v1]
Thu, 9 Jul 2026 01:18:17 UTC (2,273 KB)
— Originally published at arxiv.org
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.CL
See more →Quantifying Prior Dominance in Systems
The study introduces the Normalized Context Utilization (NCU) metric to evaluate Retrieval-Augmented Generation (RAG) systems, revealing that Small Language Models (SLMs) outperform larger models in factual extraction. The findings indicate that traditional scaling laws yield diminishing returns, with a commercial API frequently failing against adversarial evidence due to systemic confidence collapse.