Towards A Cultural Intelligence and Values Inferences Quality Benchmark for Community Values and Common Knowledge

TL;DR

This paper introduces CIVIQ, a benchmark for evaluating the cultural alignment of large language models with community values and common knowledge, addressing the gap in culturally-informed AI evaluation tools.

Contribution

It adapts the Korean KorNAT benchmark development process to create CIVIQ, a new tool for assessing cultural intelligence and values inference in LLMs within the US context.

Findings

CIVIQ provides a structured way to evaluate cultural alignment in LLMs.

The benchmark highlights gaps in current models' understanding of diverse community values.

It lays the groundwork for future research on culturally-aware AI systems.

Abstract

Large language models (LLMs) have emerged as a powerful technology, and thus, we have seen widespread adoption and use on software engineering teams. Most often, LLMs are designed as "general purpose" technologies meant to represent the general population. Unfortunately, this often means alignment with predominantly Western Caucasian narratives and misalignment with other cultures and populations that engage in collaborative innovation. In response to this misalignment, there have been recent efforts centered on the development of "culturally-informed" LLMs, such as ChatBlackGPT, that are capable of better aligning with historically marginalized experiences and perspectives. Despite this progress, there has been little effort aimed at supporting our ability to develop and evaluate culturally-informed LLMs. A recent effort proposed an approach for developing a national alignment benchmark that emphasizes alignment with national social values and common knowledge. However, given the range of cultural identities present in the United States (U.S.), a national alignment benchmark is an ineffective goal for broader representation. To help fill this gap in this US context, we propose a replication study that translates the process used to develop KorNAT, a Korean National LLM alignment benchmark, to develop CIVIQ, a Cultural Intelligence and Values Inference Quality benchmark centered on alignment with community social values and common knowledge. Our work provides a critical foundation for research and development aimed at cultural alignment of AI technologies in practice.