mlsanityData to 14 Oct 2026

The Need for a Socially-Grounded Persona Framework for User Simulation

arXiv 2601.07110 ↗First posted 12 Jan 2026cs.CL cs.AI cs.CYcited by 15

Not on the map. Why?

Abstract (arXiv)

Synthetic personas are widely used to condition large language models (LLMs) for social simulation, yet most personas are still constructed from coarse sociodemographic attributes or summaries. We revisit persona creation by introducing SCOPE, a socially grounded framework for persona construction and evaluation, built from a 141-item, two-hour sociopsychological protocol collected from 124 U.S.-based participants. Across seven models, we find that demographic-only personas are a structural bottleneck: demographics explain only ~1.5% of variance in human response similarity. Adding sociopsychological facets improves behavioral prediction and reduces over-accentuation, and non-demographic personas based on values and identity achieve strong alignment with substantially lower bias. These trends generalize to SimBench (441 aligned questions), where SCOPE personas outperform default prompting and NVIDIA Nemotron personas, and SCOPE augmentation improves Nemotron-based personas. Our results indicate that persona quality depends on sociopsychological structure rather than demographic templates or summaries.

Key facts. Cited by 15 papers in the map (data through 14 Oct 2026). Builds on 19 of its 49 references that are in the map. First posted on arXiv 12 Jan 2026.

Builds on

19 of 49 references.

References · 19

  1. 2023Generative Agents: Interactive Simulacra of Human Behaviorcited by 3,568
  2. 2018Personalizing Dialogue Agents: I have a dog, do you have pets too?cited by 828
  3. 2022Out of One, Many: Using Language Models to Simulate Human Samplescited by 752
  4. 2024Scaling Synthetic Data Creation with 1,000,000,000 Personascited by 417
  5. 2024LLM Agents Grounded in Self-Reports Enable General-Purpose Simulation of Individualscited by 249
  6. 2023Marked Personas: Using Natural Language Prompts to Measure Stereotypes in Language Modelscited by 207
  7. 2023Bias Runs Deep: Implicit Reasoning Biases in Persona-Assigned LLMscited by 199
  8. 2023Art or Artifice? Large Language Models and the False Promise of Creativitycited by 172
  9. 2023The Challenge of Using LLMs to Simulate Human Behavior: A Causal Inference Perspectivecited by 59
  10. 2025SimBench: Benchmarking the Ability of Large Language Models to Simulate Human Behaviorscited by 52
  11. 2025Finetuning LLMs for Human Behavior Prediction in Social Science Experimentscited by 39
  12. 2025Whose Personae? Synthetic Persona Experiments in LLM Research and Pathways to Transparencycited by 14
7 more
  1. 2025DeepTRACE: Auditing Deep Research AI Systems for Tracking Reliability Across Citations and Evidencecited by 10
  2. 2025Improving Language Model Personas via Rationalization with Psychological Scaffoldscited by 9
  3. 2025SPeCtrum: A Grounded Framework for Multidimensional Identity Representation in LLM-Based Agentcited by 8
  4. 2025A Tale of Two Identities: An Ethical Audit of Human and AI-Crafted Personascited by 8
  5. 2025ValueSim: Generating Backstories to Model Individual Value Systemscited by 4
  6. 2023We are what we repeatedly do: Inducing and deploying habitual schemas in persona-based responsescited by 3
  7. 2025Seeing is Believing? Evaluating Vision-Language Model Susceptibility in Agent-to-Agent Multimodal Persuasioncited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 15.

15 papers.

Sort:

By area:

  1. 2026One Persona, Many Cues, Different Results: How Sociodemographic Cues Impact LLM Personalizationcited by 9
  2. 2026Different Demographic Cues Yield Inconsistent Conclusions About LLM Personalization and Biascited by 4
  3. 2026Stories of Your Life as Others: A Round-Trip Evaluation of LLM-Generated Life Stories Conditioned on Rich Psychometric Profilescited by 2
  4. 2026Simulating Human Memory with Language Modelscited by 2
  5. 2026Topics as Proxies for Sociodemographics: How Conversational Context Affects LLM Answerscited by 2
  6. 2026CollabSim: A CSCW-Grounded Methodology for Investigating Collaborative Competence of LLM Agents through Controlled Multi-Agent Experimentscited by 2
  7. 2026Best Friends, Not Forever: Evaluating Long-Horizon Persona Collapse and Behavioral Drift in AI Companionscited by 2
  8. 2026Democratic ICAI: Debating Our Way to Steering Principles from Preferencescited by 1
  9. 2026Large Behavior Model: A Promptable Digital Twin of the Retail Customercited by 1
  10. 2026PHASE-Tree: Modeling Character-State Evolution in Long-Horizon Role-Playing Dialoguecited by 1
  11. 2026InterviewSim: A Scalable Framework for Interview-Grounded Personality Simulationcited by 0
  12. 2026The textitSilicon Society Cookbook: Design Space of LLM-based Social Simulationscited by 0
3 more
  1. 2026PrivacySIM: Evaluating LLM Simulation of User Privacy Behaviorcited by 0
  2. 2026ALSO: Adversarial Online Strategy Optimization for Social Agentscited by 0
  3. 2026RealCompanion: Benchmarking Human Understanding from Reasoning over Longitudinal Real-World Conversationscited by 0