mlsanityData to 7 Oct 2026

Benchmarking Distributional Alignment of Large Language Models

arXiv 2411.05403 ↗First posted 8 Nov 2024cs.CL cs.AIcited by 58

On the map: LLM alignment / RLHF / preference optimisation → Human values and AI feedback alignment

Abstract (arXiv)

Language models (LMs) are increasingly used as simulacra for people, yet their ability to match the distribution of views of a specific demographic group and be textitdistributionally aligned remains uncertain. This notion of distributional alignment is complex, as there is significant variation in the types of attributes that are simulated. Prior works have underexplored the role of three critical variables -- the question domain, steering method, and distribution expression method -- which motivates our contribution of a benchmark explicitly addressing these dimensions. We construct a dataset expanding beyond political values, create human baselines for this task, and evaluate the extent to which an LM can align with a particular group's opinion distribution to inform design choices of such simulation systems. Our analysis reveals open problems regarding if, and how, LMs can be used to simulate humans, and that LLMs can more accurately describe the opinion distribution than simulate such distributions.

Key facts. Cited by 58 papers in the map (data through 7 Oct 2026). Placed in Human values and AI feedback alignment, one of 10 directions in LLM alignment / RLHF / preference optimisation. Builds on 27 of its 81 references that are in the map. First posted on arXiv 8 Nov 2024.

Builds on

27 of 81 references.

Closest ancestors · 3

Same direction, newest first

  1. 2024Modular Pluralism: Pluralistic Alignment via Multi-LLM Collaboration22 Jun
  2. 2024The PRISM Alignment Dataset: What Participatory, Representative and Individualised Human Feedback Reveals About the Subjective and Multicultural Alignment of Large Language Models24 Apr
  3. 2024A Roadmap to Pluralistic Alignment7 Feb

Other references · 24

  1. 2022Language Models (Mostly) Know What They Knowcited by 1,783
  2. 2024Refusal in Language Models Is Mediated by a Single Directioncited by 960
  3. 2023Steering Language Models With Activation Engineeringcited by 824
  4. 2024Can LLMs Generate Novel Research Ideas? A Large-Scale Human Study with 100+ NLP Researcherscited by 348
  5. 2024Large language models that replace human participants can harmfully misportray and flatten identity groupscited by 137
  6. 2023Cultural Alignment in Large Language Models: An Explanatory Analysis Based on Hofstede's Cultural Dimensionscited by 111
  7. 2024PersonaGym: Evaluating Persona Agents and LLMscited by 89
  8. 2023Questioning the Survey Responses of Large Language Modelscited by 70
  9. 2024Beyond prompt brittleness: Evaluating the reliability and consistency of political worldviews in LLMscited by 55
  10. 2024LLM Processes: Numerical Predictive Distributions Conditioned on Natural Languagecited by 54
  11. 2024PAL: Pluralistic Alignment Framework for Learning from Heterogeneous Preferencescited by 53
  12. 2023Large Language Models as Superpositions of Cultural Perspectivescited by 48
12 more
  1. 2023Diminished Diversity-of-Thought in a Standard Large Language Modelcited by 43
  2. 2024Beyond Demographics: Aligning Role-playing LLM-based Agents Using Human Belief Networkscited by 39
  3. 2024Few-shot Personalization of LLMs with Mis-aligned Responsescited by 30
  4. 2024What Are the Odds? Language Models Are Capable of Probabilistic Reasoningcited by 25
  5. 2024Heads, Tails, and AI Fails: LLMs, Randomness, and Human Judgmentscited by 23
  6. 2024CIVICS: Building a Dataset for Examining Culturally-Informed Values in Large Language Modelscited by 16
  7. 2024Is Self-knowledge and Action Consistent or Not: Investigating Large Language Model's Personalitycited by 8
  8. 2024Are Large Language Models Chameleons? An Attempt to Simulate Social Surveyscited by 8
  9. 2024Using LLMs to Model the Beliefs and Preferences of Targeted Populationscited by 6
  10. 2024Aligning Large Language Models with Diverse Political Viewpointscited by 6
  11. 2024Recovering Mental Representations from Large Language Models with Markov Chain Monte Carlocited by 4
  12. 2024Implicit Probabilistic Reasoning Does Not Reflect Explicit Answers in Large Language Modelscited by 3

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 21; 2026: 37.

58 papers.

Sort:

By area:

  1. 2025LLM Social Simulations Are a Promising Research Methodcited by 97
  2. 2025Language Model Fine-Tuning on Scaled Survey Data for Predicting Distributions of Public Opinionscited by 64
  3. 2025Base Models Beat Aligned Models at Randomness and Creativitycited by 48
  4. 2025SimBench: Benchmarking the Ability of Large Language Models to Simulate Human Behaviorscited by 48
  5. 2025Beyond Demographics: Fine-tuning Large Language Models to Predict Individuals' Subjective Text Perceptionscited by 31
  6. 2026HumanLM: Simulating Users with State Alignment Beats Response Imitationcited by 31
  7. 2026This human study did not involve human subjects: Validating LLM simulations as behavioral evidencecited by 26
  8. 2025FSPO: Few-Shot Optimization of Synthetic Preferences Personalizes to Real Userscited by 14
  9. 2025Can Reasoning Help Large Language Models Capture Human Annotator Disagreement?cited by 9
  10. 2025Revisiting LLM Value Probing Strategies: Are They Robust and Expressive?cited by 8
  11. 2025Survey Response Generation: Generating Closed-Ended Survey Responses In-Silico with Large Language Modelscited by 8
  12. 2025Flipping Against All Odds: Reducing LLM Coin Flip Bias via Verbalized Rejection Samplingcited by 7
38 more
  1. 2025The Pluralistic Moral Gap: Understanding Judgment and Value Differences between Humans and Large Language Modelscited by 7
  2. 2026COLD-Steer: Steering Large Language Models via In-Context One-step Learning Dynamicscited by 7
  3. 2025Using LLMs to Advance the Cognitive Science of Collectivescited by 6
  4. 2026Evaluating Alignment of Behavioral Dispositions in LLMscited by 6
  5. 2025Distribution Shift Alignment Helps LLMs Simulate Survey Response Distributionscited by 5
  6. 2025A large-scale evaluation of commonsense knowledge in humans and large language modelscited by 4
  7. 2026The Illusion of Intervention: Your LLM-Simulated Experiment is an Observational Studycited by 4
  8. 2025Improving the Distributional Alignment of LLMs using Supervisioncited by 2
  9. 2026InfiCoEvalChain: A Blockchain-Based Decentralized Framework for Collaborative LLM Evaluationcited by 2
  10. 2026No Single Best Model for Diversity: Learning a Router for Sample Diversitycited by 2
  11. 2026Structuring the Space of Sociotechnical Alignmentcited by 2
  12. 2026Partition, Prompt, Aggregate: Statistical Self-Consistency in Language Modelscited by 2
  13. 2025Express Your Doubts -- Probabilistic World Modeling Should not be Based on Token logprobscited by 1
  14. 2025Too Open for Opinion? Embracing Open-Endedness in Large Language Models for Social Simulationcited by 1
  15. 2025QSTN: A Modular Framework for Robust Questionnaire Inference with Large Language Modelscited by 1
  16. 2026CommunityBench: Benchmarking Community-Level Alignment across Diverse Groups and Taskscited by 1
  17. 2026LLM-SAA: LLM-persona Generated Distributions for Decision-makingcited by 1
  18. 2026From Fallback to Frontline: When Can LLMs be Superior Annotators of Human Perspectives?cited by 1
  19. 2026In-Context Learning for the Imputation of Public Opinion Data with Large Language Modelscited by 1
  20. 2026Beyond Value Benchmarks: Measuring Value-Structure Alignment in Large Language Models via Symmetric Q-Sortscited by 1
  21. 2026Instruction-Tuned Language Models Cannot Sample from Distributions They Can Describecited by 1
  22. 2025Balancing Quality and Variation: Spam Filtering Distorts Data Label Distributionscited by 0
  23. 2025DeMeVa at LeWiDi-2025: Modeling Perspectives with In-Context Learning and Label Distribution Learningcited by 0
  24. 2025FlockVote: LLM-Empowered Agent-Based Modeling for Simulating U.S. Presidential Electionscited by 0
  25. 2026Evidence-based Distributional Alignment for Large Language Modelscited by 0
  26. 2026RADIUS: Ranking, Distribution, and Significance - A Comprehensive Alignment Suite for Survey Simulationcited by 0
  27. 2026Not All Subjectivity Is the Same! Defining Desiderata for the Evaluation of Subjectivity in NLPcited by 0
  28. 2026Beyond Facts: Benchmarking Distributional Reading Comprehension in Large Language Modelscited by 0
  29. 2026Predicting Causal Effects from Natural Language Queries using Structured Representationscited by 0
  30. 2026KnowledgeGain: Evaluating and Optimizing Science News Generation for Reader Learningcited by 0
  31. 2026SHALA-LLM: Smartly Handling Ambiguous Labels in Aligning LLMscited by 0
  32. 2026Evaluating Pluralism in LLMs through Latent Perspectivescited by 0
  33. 2026BehaviorBench: Benchmarking Foundation Models for Behavioral Science Taskscited by 0
  34. 2026Video2Reaction: Mapping Video to Audience Reaction Distribution in the Wildcited by 0
  35. 2026A Roadmap to Impactful Pluralistic Alignment Researchcited by 0
  36. 2026PALMs: Using Multi Construct-Grounded Rationales for Modeling Population Preferences in LLMscited by 0
  37. 2026Emulate or Estimate? The Divergent Strengths of Base and Post-Trained Language Models for Opinion Simulationcited by 0
  38. 2026Where Models Converge and Humans Diverge: A Coverage Framework for Distributional Pluralism in Open-Ended Generationcited by 0

50 most cited of 58 All 58 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2026The Limits of Simulated Societies: How Post-Training and Survey Fine-Tuning Erase Cross-Cultural Variance
  2. 2025On the Alignment of Large Language Models with Global Human Opinion
  3. 2026Controlling Distributional Bias in Multi-Round LLM Generation via KL-Optimized Fine-Tuning
  4. 2024Psychometric Alignment: Capturing Human Knowledge Distributions via Language Models
  5. 2025Multilingual != Multicultural: Evaluating Gaps Between Multilingual Capabilities and Cultural Alignment in LLMs
  6. 2025Randomness, Not Representation: The Unreliability of Evaluating Cultural Alignment in LLMs
  7. 2025DiverValue-Bench: A Benchmark and Fine-Tuning Framework for Aligning Large Language Models with Diverse Human Values
  8. 2026Beyond Marginal Distributions: A Framework to Evaluate the Representativeness of Demographic-Aligned LLMs