mlsanityData to 14 Oct 2026

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

arXiv 2606.28963 ↗First posted 27 Jun 2026cs.CL cs.CY cs.LGcited by 3

Not on the map. Why?

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.

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

20 of 29 references.

References · 20

  1. 2025Qwen3 Technical Reportcited by 8,820
  2. 2022Out of One, Many: Using Language Models to Simulate Human Samplescited by 752
  3. 2023Prediction-Powered Inferencecited by 220
  4. 2023Do LLMs exhibit human-like response biases? A case study in survey designcited by 180
  5. 2024Large language models that replace human participants can harmfully misportray and flatten identity groupscited by 144
  6. 2025Specializing Large Language Models to Simulate Survey Response Distributions for Global Populationscited by 53
  7. 2024Random Silicon Sampling: Simulating Human Sub-Population Opinion Using a Large Language Model Based on Group-Level Demographic Informationcited by 49
  8. 2025Large Language Model Psychometrics: A Systematic Review of Evaluation, Validation, and Enhancementcited by 45
  9. 2025Finetuning LLMs for Human Behavior Prediction in Social Science Experimentscited by 39
  10. 2025Valid Survey Simulations with Limited Human Data: The Roles of Prompting, Fine-Tuning, and Rectificationcited by 18
  11. 2025The threat of analytic flexibility in using large language models to simulate human datacited by 15
  12. 2025Prompt Perturbations Reveal Human-Like Biases in Large Language Model Survey Responsescited by 13
8 more
  1. 2025LLMs Reproduce Human Purchase Intent via Semantic Similarity Elicitation of Likert Ratingscited by 12
  2. 2024LLM-Mirror: A Generated-Persona Approach for Survey Pre-Testingcited by 7
  3. 2025Distribution Shift Alignment Helps LLMs Simulate Survey Response Distributionscited by 5
  4. 2026Overstating Attitudes, Ignoring Networks: LLM Biases in Simulating Misinformation Susceptibilitycited by 5
  5. 2025ChatGPT is not A Man but Das Man: Representativeness and Structural Consistency of Silicon Samples Generated by Large Language Modelscited by 4
  6. 2026Restoring Heterogeneity in LLM-based Social Simulation: An Audience Segmentation Approachcited by 4
  7. 2024ChatGPT vs Social Surveys: Probing Objective and Subjective Silicon Populationcited by 3
  8. 2025Mitigating Social Desirability Bias in Random Silicon Samplingcited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 3.

3 papers.

  1. 2026Marginal Fidelity Does Not Establish User Simulation in Demographic Synthetic Survey Panels: Response Contracts, Support Collapse and Conditioning Failurecited by 0
  2. 2026Cultural Divergence Preservation: Diagnosing Flattening and Caricature in LLM-Simulated Survey Populationscited by 0
  3. 2026Shared Autoregressive Context Can Distort Relationships in Synthetic Datacited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Can Finetuing LLMs on Small Human Samples Increase Heterogeneity, Alignment, and Belief-Action Coherence?
  2. 2025In Silico Development of Psychometric Scales: Feasibility of Representative Population Data Simulation with LLMs
  3. 2026Measurement Validity in LLM Cultural Alignment
  4. 2026The Limits of Simulated Societies: How Post-Training and Survey Fine-Tuning Erase Cross-Cultural Variance
  5. 2026Silicon Sampling via Cross-Survey Transfer
  6. 2024Improving and Assessing the Fidelity of Large Language Models Alignment to Online Communities
  7. 2026Evaluating LLMs as Human Surrogates in Controlled Experiments
  8. 2026On the Credibility of Evaluating LLMs using Survey Questions