# Beyond Cultural Persona Bias: Response Geometry as a Measure for Language Models and People

arXiv: https://arxiv.org/abs/2607.02368 · HTML: https://mlsanity.com/paper/2607.02368/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 14 of its 27 references that are in the map.
- First posted on arXiv 2 Jul 2026.

## Builds on (most cited first, 14 of 14)

- [UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction](https://mlsanity.com/paper/1802.03426/) (2018)
- [Calibrate Before Use: Improving Few-Shot Performance of Language Models](https://mlsanity.com/paper/2102.09690/) (2021)
- [Out of One, Many: Using Language Models to Simulate Human Samples](https://mlsanity.com/paper/2209.06899/) (2022)
- [Personality Traits in Large Language Models](https://mlsanity.com/paper/2307.00184/) (2023)
- [Artificial Hivemind: The Open-Ended Homogeneity of Language Models (and Beyond)](https://mlsanity.com/paper/2510.22954/) (2025)
- [Evaluating and Inducing Personality in Pre-trained Language Models](https://mlsanity.com/paper/2206.07550/) (2022)
- [Questioning the Survey Responses of Large Language Models](https://mlsanity.com/paper/2306.07951/) (2023)
- [Do LLMs Have Distinct and Consistent Personality? TRAIT: Personality Testset designed for LLMs with Psychometrics](https://mlsanity.com/paper/2406.14703/) (2024)
- [The Prompt Makes the Person(a): A Systematic Evaluation of Sociodemographic Persona Prompting for Large Language Models](https://mlsanity.com/paper/2507.16076/) (2025)
- [Self-Assessment Tests are Unreliable Measures of LLM Personality](https://mlsanity.com/paper/2309.08163/) (2023)
- [Challenging the Validity of Personality Tests for Large Language Models](https://mlsanity.com/paper/2311.05297/) (2023)
- [Robustness in Large Language Models: A Survey of Mitigation Strategies and Evaluation Metrics](https://mlsanity.com/paper/2505.18658/) (2025)
- [Can LLMs Express Personality Across Cultures? Introducing CulturalPersonas for Evaluating Trait Alignment](https://mlsanity.com/paper/2506.05670/) (2025)
- [ConsistencyAI: A Benchmark to Assess LLMs' Factual Consistency When Responding to Different Demographic Groups](https://mlsanity.com/paper/2510.13852/) (2025)

## Abstract (arXiv)

Researchers are increasingly prompting language models with cultural personas so that they answer questionnaires as diverse human groups would. However, studies often report that these persona responses do not predict the models' behavior. We argue that whether the persona is Western or non-Western, the model's answers carry a cultural preference, and that this preference masks the variables that could predict its behavior. We examine the prompted preference at two levels: aggregate Big Five scores and response geometry, the correlation matrix between dimensions within each respondent, represented as a symmetric positive-definite (SPD) matrix. In DeepSeek-V3 and GPT-4o, when the item order is fixed, the prompted preference appears at both levels. Randomizing item order reduces it in the response geometry, while leaving it in the Big Five scores: the Big Five still separate the two personas at 95--97%, while response geometry falls to chance. In randomized administration, the response geometry still contains information about the respondent. In DeepSeek-V4-Flash, SPD features are associated with whether the same instance sorts Asian American public figures as ``foreign'' in a text IAT (p = .047), while the Big Five do not. In human studies conducted using randomized item order, the response geometry identifies people from test-retest sessions in 81 -- 85%, against chance 0.4%, and predicts sex and, weakly, age. In addition, SPD features also predict complex human behavior (e.g., cumulative GPA, R2 = 0.28, permutation p = .003), where the Big Five do not.