mlsanityData to 7 Oct 2026

Questioning the Survey Responses of Large Language Models

arXiv 2306.07951 ↗First posted 13 Jun 2023cs.CLcited by 70

Not on the map. Why?

Abstract (arXiv)

Surveys have recently gained popularity as a tool to study large language models. By comparing survey responses of models to those of human reference populations, researchers aim to infer the demographics, political opinions, or values best represented by current language models. In this work, we critically examine this methodology on the basis of the well-established American Community Survey by the U.S. Census Bureau. Evaluating 43 different language models using de-facto standard prompting methodologies, we establish two dominant patterns. First, models' responses are governed by ordering and labeling biases, for example, towards survey responses labeled with the letter "A". Second, when adjusting for these systematic biases through randomized answer ordering, models across the board trend towards uniformly random survey responses, irrespective of model size or pre-training data. As a result, in contrast to conjectures from prior work, survey-derived alignment measures often permit a simple explanation: models consistently appear to better represent subgroups whose aggregate statistics are closest to uniform for any survey under consideration.

Key facts. Cited by 70 papers in the map (data through 7 Oct 2026). Builds on 31 of its 74 references that are in the map. First posted on arXiv 13 Jun 2023.

Builds on

31 of 74 references.

References · 31

  1. 2020Language Models are Few-Shot Learnerscited by 32,199
  2. 2023GPT-4 Technical Reportcited by 18,132
  3. 2022Training language models to follow instructions with human feedbackcited by 15,712
  4. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 13,958
  5. 2020Measuring Massive Multitask Language Understandingcited by 7,768
  6. 2016SQuAD: 100,000+ Questions for Machine Comprehension of Textcited by 4,115
  7. 2018Can a Suit of Armor Conduct Electricity? A New Dataset for Open Book Question Answeringcited by 2,114
  8. 2018CommonsenseQA: A Question Answering Challenge Targeting Commonsense Knowledgecited by 1,793
  9. 2023Pythia: A Suite for Analyzing Large Language Models Across Training and Scalingcited by 1,734
  10. 2021Calibrate Before Use: Improving Few-Shot Performance of Language Modelscited by 1,371
  11. 2021Fantastically Ordered Prompts and Where to Find Them: Overcoming Few-Shot Prompt Order Sensitivitycited by 1,122
  12. 2019How Can We Know What Language Models Know?cited by 953
19 more
  1. 2022Discovering Language Model Behaviors with Model-Written Evaluationscited by 805
  2. 2023Whose Opinions Do Language Models Reflect?cited by 759
  3. 2022Out of One, Many: Using Language Models to Simulate Human Samplescited by 737
  4. 2020How Much Knowledge Can You Pack Into the Parameters of a Language Model?cited by 713
  5. 2022GPT-NeoX-20B: An Open-Source Autoregressive Language Modelcited by 711
  6. 2022Using Large Language Models to Simulate Multiple Humans and Replicate Human Subject Studiescited by 506
  7. 2021Retiring Adult: New Datasets for Fair Machine Learningcited by 443
  8. 2021Persistent Anti-Muslim Bias in Large Language Modelscited by 376
  9. 2023From Pretraining Data to Language Models to Downstream Tasks: Tracking the Trails of Political Biases Leading to Unfair NLP Modelscited by 257
  10. 2023Can Large Language Models Transform Computational Social Science?cited by 245
  11. 2022Leveraging Large Language Models for Multiple Choice Question Answeringcited by 199
  12. 2015A Neural Conversational Modelcited by 195
  13. 2023The political ideology of conversational AI: Converging evidence on ChatGPT's pro-environmental, left-libertarian orientationcited by 183
  14. 2021Reframing Instructional Prompts to GPTk's Languagecited by 176
  15. 2020UnQovering Stereotyping Biases via Underspecified Questionscited by 124
  16. 2023The Self-Perception and Political Biases of ChatGPTcited by 66
  17. 2023Generative Table Pre-training Empowers Models for Tabular Predictioncited by 55
  18. 2022CommunityLM: Probing Partisan Worldviews from Language Modelscited by 36
  19. 2022FairGBM: Gradient Boosting with Fairness Constraintscited by 25

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 1; 2024: 15; 2025: 23; 2026: 31.

70 papers.

Sort:

By area:

  1. 2023Challenges and Applications of Large Language Modelscited by 251
  2. 2024Personalized Language Modeling from Personalized Human Feedbackcited by 124
  3. 2025LLM Social Simulations Are a Promising Research Methodcited by 97
  4. 2025Language Model Fine-Tuning on Scaled Survey Data for Predicting Distributions of Public Opinionscited by 64
  5. 2024Eliminating Position Bias of Language Models: A Mechanistic Approachcited by 61
  6. 2024Benchmarking Distributional Alignment of Large Language Modelscited by 58
  7. 2024Beyond prompt brittleness: Evaluating the reliability and consistency of political worldviews in LLMscited by 55
  8. 2024Virtual Personas for Language Models via an Anthology of Backstoriescited by 49
  9. 2025SimBench: Benchmarking the Ability of Large Language Models to Simulate Human Behaviorscited by 48
  10. 2024Limits to scalable evaluation at the frontier: LLM as Judge won't beat twice the datacited by 46
  11. 2024Evaluating Cultural Adaptability of a Large Language Model via Simulation of Synthetic Personascited by 36
  12. 2025The Personality Illusion: Revealing Dissociation Between Self-Reports & Behavior in LLMscited by 35
38 more
  1. 2024Training on the Test Task Confounds Evaluation and Emergencecited by 33
  2. 2026This human study did not involve human subjects: Validating LLM simulations as behavioral evidencecited by 26
  3. 2024Evaluating language models as risk scorescited by 24
  4. 2024Toward LLM-Agent-Based Modeling of Transportation Systems: A Conceptual Frameworkcited by 19
  5. 2024Look at the Text: Instruction-Tuned Language Models are More Robust Multiple Choice Selectors than You Thinkcited by 13
  6. 2024Vox Populi, Vox AI? Using Language Models to Estimate German Public Opinioncited by 13
  7. 2024Measuring Human and AI Values Based on Generative Psychometrics with Large Language Modelscited by 13
  8. 2025A Large Scale Analysis of Gender Biases in Text-to-Image Generative Modelscited by 12
  9. 2025SATA-BENCH: Select All That Apply Benchmark for Multiple Choice Questionscited by 9
  10. 2025Position: Stop Evaluating AI with Human Tests, Develop Principled, AI-specific Tests insteadcited by 7
  11. 2025Valid Inference with Imperfect Synthetic Datacited by 7
  12. 2024LABOR-LLM: Language-Based Occupational Representations with Large Language Modelscited by 5
  13. 2024United in Diversity? Contextual Biases in LLM-Based Predictions of the 2024 European Parliament Electionscited by 5
  14. 2025Generative Psycho-Lexical Approach for Constructing Value Systems in Large Language Modelscited by 4
  15. 2025Human Preferences in Large Language Model Latent Space: A Technical Analysis on the Reliability of Synthetic Data in Voting Outcome Predictioncited by 4
  16. 2025Person-Centric Annotations of LAION-400M: Auditing Bias and Its Transfer to Modelscited by 4
  17. 2025Prompting for Policy: Forecasting Macroeconomic Scenarios with Synthetic LLM Personascited by 4
  18. 2025Localizing Persona Representations in LLMscited by 3
  19. 2025On the Inevitability of Left-Leaning Political Bias in Aligned Language Modelscited by 3
  20. 2026When Synthetic Users Fail: A Cross-Domain Benchmark of LLM-Simulated Human Survey Responsescited by 3
  21. 2025From Stability to Inconsistency: A Study of Moral Preferences in LLMscited by 2
  22. 2025Epidemiology of Large Language Models: A Benchmark for Observational Distribution Knowledgecited by 2
  23. 2026HumanStudy-Bench: Towards AI Agent Design for Participant Simulationcited by 2
  24. 2026When Can Digital Personas Reliably Approximate Human Survey Findings?cited by 2
  25. 2026Preregistration for Experiments with AI Agentscited by 2
  26. 2026Partition, Prompt, Aggregate: Statistical Self-Consistency in Language Modelscited by 2
  27. 2025NeuroGen: Neural Network Parameter Generation via Large Language Modelscited by 1
  28. 2025Hypothesis Testing for Quantifying LLM-Human Misalignment in Multiple Choice Settingscited by 1
  29. 2025Positional Bias in Binary Question Answering: How Uncertainty Shapes Model Preferencescited by 1
  30. 2025QuEst: Enhancing Estimates of Quantile-Based Distributional Measures Using Model Predictionscited by 1
  31. 2025Speciesism in AI: Evaluating Discrimination Against Animals in Large Language Modelscited by 1
  32. 2025LLM Personas as a Substitute for Field Experiments in Method Benchmarkingcited by 1
  33. 2026Distributional Open-Ended Evaluation of LLM Cultural Value Alignment Based on Value Codebookcited by 1
  34. 2026Do Large Language Models Plan Answer Positions? Position Bias in Multiple-Choice Question Generationcited by 1
  35. 2026Consequential Behaviour and Representational Fairness in the Validation of Synthetic Researchcited by 1
  36. 2025How Large Language Models Systematically Misrepresent American Climate Opinionscited by 0
  37. 2026Characterizing the ability of LLMs to recapitulate Americans' distributional responses to public opinion polling questions across political issuescited by 0
  38. 2026Jump Start or False Start? A Theoretical and Empirical Evaluation of LLM-initialized Banditscited by 0

50 most cited of 70 All 70 (JSON)