Can Persona-Prompted LLMs Emulate Subgroup Values? An Empirical Analysis of Generalisability and Fairness in Cultural Alignment

• URL: http://arxiv.org/abs/2604.12851v1
• Date: Tue, 14 Apr 2026 15:06:13 GMT
• Title: Can Persona-Prompted LLMs Emulate Subgroup Values? An Empirical Analysis of Generalisability and Fairness in Cultural Alignment
• Abstract summary: We show that even state-of-the-art models like GPT-4.1 achieve only 57.4% accuracy in predicting modal preferences.<n>We find significant pre-existing performance biases, where models better emulate young, male, Chinese, and Christian personas.
• Score: 75.88231994139132
• License: http://creativecommons.org/licenses/by-sa/4.0/
• Abstract: Despite their global prevalence, many Large Language Models (LLMs) are aligned to a monolithic, often Western-centric set of values. This paper investigates the more challenging task of fine-grained value alignment: examining whether LLMs can emulate the distinct cultural values of demographic subgroups. Using Singapore as a case study and the World Values Survey (WVS), we examine the value landscape and show that even state-of-the-art models like GPT-4.1 achieve only 57.4% accuracy in predicting subgroup modal preferences. We construct a dataset of over 20,000 samples to train and evaluate a range of models. We demonstrate that simple fine-tuning on structured numerical preferences yields substantial gains, improving accuracy on unseen, out-of-distribution subgroups by an average of 17.4%. These gains partially transfer to open-ended generation. However, we find significant pre-existing performance biases, where models better emulate young, male, Chinese, and Christian personas. Furthermore, while fine-tuning improves average performance, it widens the disparity between subgroups when measured by distance-aware metrics. Our work offers insights into the limits and fairness implications of subgroup-level cultural alignment.

Related papers

• PALMs: Using Multi Construct-Grounded Rationales for Modeling Population Preferences in LLMs [10.940747084451559]
  We introduce Population Aligned Language Models (PALMs), a suite of models each aligned to specific populations.<n>PALMs are Evaluated across four dimensions: personality, values and beliefs, cultural norms, and morality.<n>We demonstrate strong generalization to downstream applications with- out task-specific supervision.
  arXiv  Detail & Related papers  (2026-08-02T19:31:59Z)
• DVMap: Fine-Grained Pluralistic Value Alignment via High-Consensus Demographic-Value Mapping [49.402442592047855]
  Current Large Language Models (LLMs) typically rely on coarse-grained national labels for pluralistic value alignment.<n>We argue that resolving this limitation requires shifting from national labels to multi-dimensional demographic constraints.<n>We propose DVMap, a framework for fine-grained pluralistic value alignment.
  arXiv  Detail & Related papers  (2026-05-14T06:06:43Z)
• Distributional Open-Ended Evaluation of LLM Cultural Value Alignment Based on Value Codebook [50.10594064510559]
  DOVE is a distributional evaluation framework that directly compares human-written text with LLM-generated outputs.<n>We show that DOVE achieves superior predictive validity, attaining a 31.56% correlation with downstream tasks, while maintaining high reliability with as few as 500 samples per culture.
  arXiv  Detail & Related papers  (2026-03-16T08:33:10Z)
• Us-vs-Them bias in Large Language Models [0.569978892646475]
  We find consistent ingroup-positive and outgroup-negative associations across foundational large language models.<n>For personas examined, conservative personas exhibit greater outgroup hostility, whereas liberal personas display stronger ingroup solidarity.
  arXiv  Detail & Related papers  (2025-12-03T07:11:22Z)
• Culture Affordance Atlas: Reconciling Object Diversity Through Functional Mapping [38.345727498425]
  Vision-Language (VL) datasets exhibit cultural biases, disproportionately favoring higher-income, Western contexts.<n>We propose a novel function-centric framework that categorizes objects by the functions they fulfill, across diverse cultural and economic contexts.
  arXiv  Detail & Related papers  (2025-12-02T19:16:39Z)
• Mind the Gaps: Auditing and Reducing Group Inequity in Large-Scale Mobility Prediction [9.369284351516358]
  Next location prediction underpins a growing number of mobility, retail, and public-health applications.<n>In this paper, we audit state-of-the-art mobility prediction models trained on a large-scale dataset.<n>We show a systematic disparity resulting from the underlying dataset, resulting in large differences in accuracy based on location and user groups.
  arXiv  Detail & Related papers  (2025-10-30T18:54:33Z)
• Exploring Cultural Variations in Moral Judgments with Large Language Models [0.5356944479760104]
  Using log-probability-based moral justifiability scores, we correlate each model's outputs with survey data covering a broad set of ethical topics.<n>Our results show that many earlier or smaller models often produce near-zero or negative correlations with human judgments.<n> advanced instruction-tuned models (including GPT-4o and GPT-4o-mini) achieve substantially higher positive correlations, suggesting they better reflect real-world moral attitudes.
  arXiv  Detail & Related papers  (2025-06-14T10:16:48Z)
• WorldPM: Scaling Human Preference Modeling [130.23230492612214]
  We propose World Preference Modeling$ (WorldPM) to emphasize this scaling potential.<n>We collect preference data from public forums covering diverse user communities.<n>We conduct extensive training using 15M-scale data across models ranging from 1.5B to 72B parameters.
  arXiv  Detail & Related papers  (2025-05-15T17:38:37Z)
• Towards Large Language Models that Benefit for All: Benchmarking Group Fairness in Reward Models [16.977176752570617]
  Large Language Models (LLMs) are increasingly powerful and accessible to human users.<n> Ensuring fairness across diverse demographic groups, i.e., group fairness, is a critical ethical concern.<n>This work benchmarks the group fairness of learned reward models.
  arXiv  Detail & Related papers  (2025-03-10T19:39:39Z)
• On the steerability of large language models toward data-driven personas [98.9138902560793]
  Large language models (LLMs) are known to generate biased responses where the opinions of certain groups and populations are underrepresented. Here, we present a novel approach to achieve controllable generation of specific viewpoints using LLMs.
  arXiv  Detail & Related papers  (2023-11-08T19:01:13Z)
• Bias and Fairness in Large Language Models: A Survey [73.87651986156006]
  We present a comprehensive survey of bias evaluation and mitigation techniques for large language models (LLMs) We first consolidate, formalize, and expand notions of social bias and fairness in natural language processing. We then unify the literature by proposing three intuitive, two for bias evaluation, and one for mitigation.
  arXiv  Detail & Related papers  (2023-09-02T00:32:55Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.