# Parametric Social Identity Injection and Diversification in Public Opinion Simulation

arXiv: https://arxiv.org/abs/2603.16142 · HTML: https://mlsanity.com/paper/2603.16142/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 26 of its 44 references that are in the map.
- First posted on arXiv 17 Mar 2026.

## Builds on (most cited first, 26 of 26)

- [LLaMA: Open and Efficient Foundation Language Models](https://mlsanity.com/paper/2302.13971/) (2023)
- [Qwen Technical Report](https://mlsanity.com/paper/2309.16609/) (2023)
- [Whose Opinions Do Language Models Reflect?](https://mlsanity.com/paper/2303.17548/) (2023)
- [Out of One, Many: Using Language Models to Simulate Human Samples](https://mlsanity.com/paper/2209.06899/) (2022)
- [Persona Vectors: Monitoring and Controlling Character Traits in Language Models](https://mlsanity.com/paper/2507.21509/) (2025)
- [CulturaX: A Cleaned, Enormous, and Multilingual Dataset for Large Language Models in 167 Languages](https://mlsanity.com/paper/2309.09400/) (2023)
- [Large language models that replace human participants can harmfully misportray and flatten identity groups](https://mlsanity.com/paper/2402.01908/) (2024)
- [Increasing Diversity While Maintaining Accuracy: Text Data Generation with Large Language Models and Human Interventions](https://mlsanity.com/paper/2306.04140/) (2023)
- [A Note on Shumailov et al. (2024): `AI Models Collapse When Trained on Recursively Generated Data'](https://mlsanity.com/paper/2410.12954/) (2024)
- [Aligning Language Models to User Opinions](https://mlsanity.com/paper/2305.14929/) (2023)
- [Sensitivity, Performance, Robustness: Deconstructing the Effect of Sociodemographic Prompting](https://mlsanity.com/paper/2309.07034/) (2023)
- [Language Model Fine-Tuning on Scaled Survey Data for Predicting Distributions of Public Opinions](https://mlsanity.com/paper/2502.16761/) (2025)
- [Algorithmic Fairness Datasets: the Story so Far](https://mlsanity.com/paper/2202.01711/) (2022)
- [Improving Diversity of Demographic Representation in Large Language Models via Collective-Critiques and Self-Voting](https://mlsanity.com/paper/2310.16523/) (2023)
- [Diminished Diversity-of-Thought in a Standard Large Language Model](https://mlsanity.com/paper/2302.07267/) (2023)
- [Parametric Retrieval Augmented Generation](https://mlsanity.com/paper/2501.15915/) (2025)
- [Cultivating Pluralism In Algorithmic Monoculture: The Community Alignment Dataset](https://mlsanity.com/paper/2507.09650/) (2025)
- [How Far Can We Extract Diverse Perspectives from Large Language Models?](https://mlsanity.com/paper/2311.09799/) (2023)
- [Population-Aligned Persona Generation for LLM-based Social Simulation](https://mlsanity.com/paper/2509.10127/) (2025)
- [Are Large Language Models (LLMs) Good Social Predictors?](https://mlsanity.com/paper/2402.12620/) (2024)
- [Debiased Large Language Models Still Associate Muslims with Uniquely Violent Acts](https://mlsanity.com/paper/2208.04417/) (2022)
- [Wisdom from Diversity: Bias Mitigation Through Hybrid Human-LLM Crowds](https://mlsanity.com/paper/2505.12349/) (2025)
- [Investigating Prosocial Behavior Theory in LLM Agents under Policy-Induced Inequities](https://mlsanity.com/paper/2505.15857/) (2025)
- [Synthesizing Public Opinions with LLMs: Role Creation, Impacts, and the Future to eDemorcacy](https://mlsanity.com/paper/2504.00241/) (2025)
- [Distribution Shift Alignment Helps LLMs Simulate Survey Response Distributions](https://mlsanity.com/paper/2510.21977/) (2025)
- [HAG: Hierarchical Demographic Tree-based Agent Generation for Topic-Adaptive Simulation](https://mlsanity.com/paper/2601.05656/) (2026)

## Built on by (most cited first, 2 of 2)

- [Mitigating Identity Essentialism in LLM Agents with Longitudinal Life Trajectories](https://mlsanity.com/paper/2608.19621/) (2026)
- [Readable, Faithful, Used: Three Dissociable Properties of Demographic Identity in a Language Model](https://mlsanity.com/paper/2608.18768/) (2026)

## Abstract (arXiv)

Large language models (LLMs) have recently been adopted as synthetic agents for public opinion simulation, offering a promising alternative to costly and slow human surveys. Despite their scalability, current LLM-based simulation methods fail to capture social diversity, producing flattened inter-group differences and overly homogeneous responses across demographic groups. We identify this limitation as a Diversity Collapse phenomenon in LLM hidden representations, where distinct social identities become increasingly indistinguishable across layers. Motivated by this observation, we propose Parametric Social Identity Injection (PSII), a general framework that injects explicit, parametric representations of demographic attributes and value orientations directly into intermediate hidden states of LLMs. Unlike prompt-based persona conditioning, PSII enables fine-grained and controllable identity modulation at the representation level. Extensive experiments on the World Values Survey using multiple open-source LLMs show that PSII significantly improves distributional fidelity and diversity, reducing KL divergence to real-world survey data while enhancing overall diversity. This work provides new insights into representation-level control of LLM agents and advances scalable, diversity-aware public opinion simulation.