aiqbench analysis
This research introduces "value diversity" as a critical system-level metric for multicultural AI, demonstrating that current multi-agent systems significantly lag behind human societies in preserving cultural plurality despite high individual alignment. The study reveals that social interactions within these systems tend to drive homogenization and consensus, thereby narrowing the scope of collective decision-making. These findings highlight an urgent need to decouple diversity metrics from simple alignment checks when evaluating culturally conditioned large language models.
Beyond Alignment: Value Diversity as a Collective Property in Multicultural Agent Systems
arXiv:2606.05985v2 Announce Type: replace Abstract: Multicultural multi-agent systems are increasingly deployed in globally diverse settings, where different agents are grounded in different cultural backgrounds. Existing cultural evaluation focuses on value alignment: how closely a single agent matches a target culture. Yet alignment is a per-agent property and cannot reveal whether a system, taken as a whole, preserves the cultural plurality it is meant to represent. We propose value diversity as a system-level evaluation axis for multicultural agent systems, defined through the dissimilarity between culturally conditioned agents' responses on a shared value survey. Using the World Values Survey, we evaluate 19 cultures and 18 backbone models across a wide range of system configurations. We find that diversity is largely uncorrelated with alignment, indicating that the two capture complementary system properties, and that current multicultural agent systems fall substantially below human societies in value diversity. Mixed-backbone systems narrow this gap but do not close it, and the gap persists across culture compositions and agent scales. Social interaction further erodes diversity by driving agents toward consensus, and a participatory budgeting case study shows that this homogenization narrows the breadth of collective decision-making. Together, our results establish value diversity as a distinct evaluation axis for multicultural multi-agent systems and reveal a persistent homogenization tendency in current LLM-based societies. Our code and data are publicly available at https://github.com/iNLP-Lab/MultiAgent-Diversity.
Abstract:Multicultural multi-agent systems are increasingly deployed in globally diverse settings, where different agents are grounded in different cultural backgrounds. Existing cultural evaluation focuses on value alignment: how closely a single agent matches a target culture. Yet alignment is a per-agent property and cannot reveal whether a system, taken as a whole, preserves the cultural plurality it is meant to represent. We propose value diversity as a system-level evaluation axis for multicultural agent systems, defined through the dissimilarity between culturally conditioned agents' responses on a shared value survey. Using the World Values Survey, we evaluate 19 cultures and 18 backbone models across a wide range of system configurations. We find that diversity is largely uncorrelated with alignment, indicating that the two capture complementary system properties, and that current multicultural agent systems fall substantially below human societies in value diversity. Mixed-backbone systems narrow this gap but do not close it, and the gap persists across culture compositions and agent scales. Social interaction further erodes diversity by driving agents toward consensus, and a participatory budgeting case study shows that this homogenization narrows the breadth of collective decision-making. Together, our results establish value diversity as a distinct evaluation axis for multicultural multi-agent systems and reveal a persistent homogenization tendency in current LLM-based societies. Our code and data are publicly available at this https URL.
Computation and Language (cs.CL); Computers and Society (cs.CY)
arXiv:2606.05985 [cs.CL]
(or arXiv:2606.05985v2 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2606.05985 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 317 words
Related coverage
Beyond Alignment: Value Diversity as a Collective Property in Multicultural Agent Systems
arXiv · Jun 5, 2026
State of the Field: AI for Epistemics and Coordination
LessWrong · Oct 2, 2026

Concrete Evaluations to Investigate the OpenAI Model That Hacked Hugging Face
AI Alignment Forum · Aug 3, 2026

An unexamined cause of the OpenAI Hugging Face hacking incident: its binary performance metric
LessWrong · Sep 23, 2026