mlsanityData to 7 Oct 2026

PERSONA: A Reproducible Testbed for Pluralistic Alignment

arXiv 2407.17387 ↗First posted 24 Jul 2024cs.CLcited by 71

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

Abstract (arXiv)

The rapid advancement of language models (LMs) necessitates robust alignment with diverse user values. However, current preference optimization approaches often fail to capture the plurality of user opinions, instead reinforcing majority viewpoints and marginalizing minority perspectives. We introduce PERSONA, a reproducible test bed designed to evaluate and improve pluralistic alignment of LMs. We procedurally generate diverse user profiles from US census data, resulting in 1,586 synthetic personas with varied demographic and idiosyncratic attributes. We then generate a large-scale evaluation dataset containing 3,868 prompts and 317,200 feedback pairs obtained from our synthetic personas. Leveraging this dataset, we systematically evaluate LM capabilities in role-playing diverse users, verified through human judges, and the establishment of both a benchmark, PERSONA Bench, for pluralistic alignment approaches as well as an extensive dataset to create new and future benchmarks. The full dataset and benchmarks are available here: https://www.synthlabs.ai/research/persona.

Key facts. Cited by 71 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 37 of its 60 references that are in the map. First posted on arXiv 24 Jul 2024.

Builds on

37 of 60 references.

Closest ancestors · 5

Same direction, newest first

  1. 2024Personalized Language Modeling from Personalized Human Feedback6 Feb
  2. 2023Personalized Soups: Personalized Large Language Model Alignment via Post-hoc Parameter Merging17 Oct
  3. 2023Aligning Language Models to User Opinions24 May
  4. 2023AlpacaFarm: A Simulation Framework for Methods that Learn from Human Feedback22 May
  5. 2020Learning to summarize from human feedback2 Sep

Other references · 32

  1. 2022Training language models to follow instructions with human feedbackcited by 15,712
  2. 2023Judging LLM-as-a-Judge with MT-Bench and Chatbot Arenacited by 8,393
  3. 2023Generative Agents: Interactive Simulacra of Human Behaviorcited by 3,500
  4. 2023Lost in the Middle: How Language Models Use Long Contextscited by 3,383
  5. 2022Scaling Laws for Reward Model Overoptimizationcited by 1,056
  6. 2020Language (Technology) is Power: A Critical Survey of "Bias" in NLPcited by 951
  7. 2023Whose Opinions Do Language Models Reflect?cited by 759
  8. 2023Towards Measuring the Representation of Subjective Global Opinions in Language Modelscited by 424
  9. 2023Character-LLM: A Trainable Agent for Role-Playingcited by 335
  10. 2022Challenges and Strategies in Cross-Cultural NLPcited by 216
  11. 2024From Persona to Personalization: A Survey on Role-Playing Language Agentscited by 206
  12. 2024Scaling Laws for Reward Model Overoptimization in Direct Alignment Algorithmscited by 139
20 more
  1. 2023Distributional Preference Learning: Understanding and Accounting for Hidden Context in RLHFcited by 125
  2. 2024Large Language Models are Superpositions of All Characters: Attaining Arbitrary Role-play via Self-Alignmentcited by 104
  3. 2024STaR-GATE: Teaching Language Models to Ask Clarifying Questionscited by 99
  4. 2024Group Robust Preference Optimization in Reward-free RLHFcited by 94
  5. 2023Eliciting Human Preferences with Language Modelscited by 90
  6. 2024Measuring Political Bias in Large Language Models: What Is Said and How It Is Saidcited by 89
  7. 2024Self-Exploring Language Models: Active Preference Elicitation for Online Alignmentcited by 63
  8. 2023Group Preference Optimization: Few-Shot Alignment of Large Language Modelscited by 61
  9. 2023ChatHaruhi: Reviving Anime Character in Reality via Large Language Modelcited by 57
  10. 2024Persona-DB: Efficient Large Language Model Personalization for Response Prediction with Collaborative Data Refinementcited by 47
  11. 2023Personas as a Way to Model Truthfulness in Language Modelscited by 46
  12. 2023Not All Countries Celebrate Thanksgiving: On the Cultural Dominance in Large Language Modelscited by 42
  13. 2023Can Large Language Models Capture Dissenting Human Voices?cited by 33
  14. 2024Assessing Political Bias in Large Language Modelscited by 32
  15. 2023Active Preference Inference using Language Models and Probabilistic Reasoningcited by 29
  16. 2024Direct Preference Optimization With Unobserved Preference Heterogeneity: The Necessity of Ternary Preferencescited by 23
  17. 2024Self-Supervised Alignment with Mutual Information: Learning to Follow Principles without Preference Labelscited by 18
  18. 2023Are Language Models Worse than Humans at Following Prompts? It's Complicatedcited by 14
  19. 2024Suppressing Pink Elephants with Direct Principle Feedbackcited by 13
  20. 2023Social Contract AI: Aligning AI Assistants with Implicit Group Normscited by 9

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 7; 2025: 37; 2026: 27.

71 papers.

Sort:

By area:

  1. 2025Know Me, Respond to Me: Benchmarking LLMs for Dynamic User Profiling and Personalized Responses at Scalecited by 163
  2. 2024Generative Reward Modelscited by 148
  3. 2025LLM Generated Persona is a Promise with a Catchcited by 77
  4. 2024PersonalLLM: Tailoring LLMs to Individual Preferencescited by 56
  5. 2025PersonaLens: A Benchmark for Personalization Evaluation in Conversational AI Assistantscited by 44
  6. 2025A Survey on Personalized Alignment -- The Missing Piece for Large Language Models in Real-World Applicationscited by 40
  7. 2024Controllable Safety Alignment: Inference-Time Adaptation to Diverse Safety Requirementscited by 35
  8. 2024Can Language Models Reason about Individualistic Human Values and Preferences?cited by 34
  9. 2025LoRe: Personalizing LLMs via Low-Rank Reward Modelingcited by 34
  10. 2024Stereotype or Personalization? User Identity Biases Chatbot Recommendationscited by 33
  11. 2026HumanLM: Simulating Users with State Alignment Beats Response Imitationcited by 31
  12. 2025Language Model Personalization via Reward Factorizationcited by 27
38 more
  1. 2025Towards a Design Guideline for RPA Evaluation: A Survey of Large Language Model-Based Role-Playing Agentscited by 22
  2. 2025Persona-Augmented Benchmarking: Evaluating LLMs Across Diverse Writing Stylescited by 15
  3. 2025FSPO: Few-Shot Optimization of Synthetic Preferences Personalizes to Real Userscited by 14
  4. 2025DeepPersona: A Generative Engine for Scaling Deep Synthetic Personascited by 14
  5. 2025Pairwise Calibrated Rewards for Pluralistic Alignmentcited by 12
  6. 2024Test-Time Alignment via Hypothesis Reweightingcited by 11
  7. 2025Whose View of Safety? A Deep DIVE Dataset for Pluralistic Alignment of Text-to-Image Modelscited by 11
  8. 2025Embedding-to-Prefix: Parameter-Efficient Personalization for Pre-Trained Large Language Modelscited by 10
  9. 2026Toward Personalized LLM-Powered Agents: Foundations, Evaluation, and Future Directionscited by 10
  10. 2025When Personalization Meets Reality: A Multi-Faceted Analysis of Personalized Preference Learningcited by 9
  11. 2025Robust Multi-Objective Controlled Decoding of Large Language Modelscited by 8
  12. 2025Adaptive Elicitation of Latent Information Using Natural Languagecited by 7
  13. 2025Doubly-Robust LLM-as-a-Judge: Externally Valid Estimation with Imperfect Personascited by 7
  14. 2026COLD-Steer: Steering Large Language Models via In-Context One-step Learning Dynamicscited by 7
  15. 2026Quantifying the Utility of User Simulators for Building Collaborative LLM Assistantscited by 6
  16. 2025Strategyproof Reinforcement Learning from Human Feedbackcited by 5
  17. 2025A Review of Incorporating Psychological Theories in LLMscited by 5
  18. 2025FaST: Feature-aware Sampling and Tuning for Personalized Preference Alignment with Limited Datacited by 5
  19. 2025Network-Level Prompt and Trait Leakage in Local Research Agentscited by 5
  20. 2025Counterfactual Reasoning for Steerable Pluralistic Value Alignment of Large Language Modelscited by 5
  21. 2025Learning to Make Friends: Coaching LLM Agents toward Emergent Social Tiescited by 5
  22. 2026Positive Alignment: Artificial Intelligence for Human Flourishingcited by 5
  23. 2025Sleepless Nights, Sugary Days: Creating Synthetic Users with Health Conditions for Realistic Coaching Agent Interactionscited by 4
  24. 2025German General Social Survey Personas: A Survey-Derived Persona Prompt Collection for Population-Aligned LLM Studiescited by 4
  25. 2026Beyond Fixed Psychological Personas: State Beats Trait, but Language Models are State-Blindcited by 4
  26. 2026The Illusion of Intervention: Your LLM-Simulated Experiment is an Observational Studycited by 4
  27. 2024SubData: Bridging Heterogeneous Datasets to Enable Theory-Driven Evaluation of Political and Demographic Perspectives in LLMscited by 3
  28. 2025EVALUESTEER: Measuring Reward Model Steerability Towards Values and Preferencescited by 3
  29. 2026Persona Prompting as a Lens on LLM Social Reasoningcited by 3
  30. 2025Designing Synthetic Discussion Generation Systems: A Case Study for Online Facilitationcited by 2
  31. 2025Prompts to Proxies: Emulating Human Preferences via a Compact LLM Ensemblecited by 2
  32. 2026Persona Generators: Generating Diverse Synthetic Personas for Arbitrary Contextscited by 2
  33. 2026Structuring the Space of Sociotechnical Alignmentcited by 2
  34. 2025Is Active Persona Inference Necessary for Aligning Small Models to Personal Preferences?cited by 1
  35. 2025Benchmarking In-context Experiential Learning Through Repeated Product Recommendationscited by 1
  36. 2025MiniScope: A Least Privilege Framework for Authorizing Tool Calling Agentscited by 1
  37. 2026HAL: Inducing Human-likeness in LLMs with Alignmentcited by 1
  38. 2025Agentic Persona Control and Task State Tracking for Realistic User Simulation in Interactive Scenarioscited by 1

50 most cited of 71 All 71 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2025Whose Personae? Synthetic Persona Experiments in LLM Research and Pathways to Transparency
  2. 2025Persona-judge: Personalized Alignment of Large Language Models via Token-level Self-judgment
  3. 2025From 1,000,000 Users to Every User: Scaling Up Personalized Preference for User-level Alignment
  4. 2026When Persona Attributes Improve Population Alignment in Large Language Models
  5. 2024A Roadmap to Pluralistic Alignment
  6. 2026EpiPersona: Persona Projection and Episode Coupling for Pluralistic Preference Modeling
  7. 2025Fair-PP: A Synthetic Dataset for Aligning LLM with Personalized Preferences of Social Equity
  8. 2024PersonaGym: Evaluating Persona Agents and LLMs