# Beyond the Mean: Three-Axis Fidelity for Aligning LLM-Based Survey Simulators from Small Pilot Data

arXiv: https://arxiv.org/abs/2606.28963 · HTML: https://mlsanity.com/paper/2606.28963/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 20 of its 29 references that are in the map.
- First posted on arXiv 27 Jun 2026.

## Builds on (most cited first, 20 of 20)

- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [Out of One, Many: Using Language Models to Simulate Human Samples](https://mlsanity.com/paper/2209.06899/) (2022)
- [Prediction-Powered Inference](https://mlsanity.com/paper/2301.09633/) (2023)
- [Do LLMs exhibit human-like response biases? A case study in survey design](https://mlsanity.com/paper/2311.04076/) (2023)
- [Large language models that replace human participants can harmfully misportray and flatten identity groups](https://mlsanity.com/paper/2402.01908/) (2024)
- [Specializing Large Language Models to Simulate Survey Response Distributions for Global Populations](https://mlsanity.com/paper/2502.07068/) (2025)
- [Random Silicon Sampling: Simulating Human Sub-Population Opinion Using a Large Language Model Based on Group-Level Demographic Information](https://mlsanity.com/paper/2402.18144/) (2024)
- [Large Language Model Psychometrics: A Systematic Review of Evaluation, Validation, and Enhancement](https://mlsanity.com/paper/2505.08245/) (2025)
- [Finetuning LLMs for Human Behavior Prediction in Social Science Experiments](https://mlsanity.com/paper/2509.05830/) (2025)
- [Valid Survey Simulations with Limited Human Data: The Roles of Prompting, Fine-Tuning, and Rectification](https://mlsanity.com/paper/2510.11408/) (2025)
- [The threat of analytic flexibility in using large language models to simulate human data](https://mlsanity.com/paper/2509.13397/) (2025)
- [Prompt Perturbations Reveal Human-Like Biases in Large Language Model Survey Responses](https://mlsanity.com/paper/2507.07188/) (2025)
- [LLMs Reproduce Human Purchase Intent via Semantic Similarity Elicitation of Likert Ratings](https://mlsanity.com/paper/2510.08338/) (2025)
- [LLM-Mirror: A Generated-Persona Approach for Survey Pre-Testing](https://mlsanity.com/paper/2412.03162/) (2024)
- [Distribution Shift Alignment Helps LLMs Simulate Survey Response Distributions](https://mlsanity.com/paper/2510.21977/) (2025)
- [Overstating Attitudes, Ignoring Networks: LLM Biases in Simulating Misinformation Susceptibility](https://mlsanity.com/paper/2602.04674/) (2026)
- [ChatGPT is not A Man but Das Man: Representativeness and Structural Consistency of Silicon Samples Generated by Large Language Models](https://mlsanity.com/paper/2507.02919/) (2025)
- [Restoring Heterogeneity in LLM-based Social Simulation: An Audience Segmentation Approach](https://mlsanity.com/paper/2604.06663/) (2026)
- [ChatGPT vs Social Surveys: Probing Objective and Subjective Silicon Population](https://mlsanity.com/paper/2409.02601/) (2024)
- [Mitigating Social Desirability Bias in Random Silicon Sampling](https://mlsanity.com/paper/2512.22725/) (2025)

## Built on by (most cited first, 3 of 3)

- [Marginal Fidelity Does Not Establish User Simulation in Demographic Synthetic Survey Panels: Response Contracts, Support Collapse and Conditioning Failure](https://mlsanity.com/paper/2609.07305/) (2026)
- [Cultural Divergence Preservation: Diagnosing Flattening and Caricature in LLM-Simulated Survey Populations](https://mlsanity.com/paper/2609.29928/) (2026)
- [Shared Autoregressive Context Can Distort Relationships in Synthetic Data](https://mlsanity.com/paper/2609.32546/) (2026)

## Abstract (arXiv)

Large language models (LLMs) are increasingly used to simulate social survey responses, yet their outputs exhibit systematic biases: marginal distributions are skewed, response variance is poorly calibrated, and predictor-outcome relationships are attenuated. We ask a simple question: given a small pilot sample of human responses, can an LLM recover the statistical characteristics of a broader population? We decompose recovery along three axes: structural fidelity, marginal fidelity, and individual fidelity. Using a COVID-19 misinformation survey as a case study, we benchmark three families of approaches: prompting, rectification, and fine-tuning. The findings suggest that fine-tuning on small pilot samples offers a balanced approach for achieving multiple forms of fidelity, but the levels of such fidelity can vary across subsamples, potentially threatening pluralistic alignment.