mlsanityData to 14 Oct 2026

Simulating Society Requires Simulating Thought

arXiv 2506.06958 ↗First posted 8 Jun 2025cs.CY cs.AI cs.MAcited by 13

On the map: LLM reasoning → Reasoning for agents, planning and code

also in 1 more
Abstract (arXiv)

Simulating society with large language models (LLMs), we argue, requires more than generating plausible behavior; it demands cognitively grounded reasoning that is structured, revisable, and traceable. LLM-based agents are increasingly used to emulate individual and group behavior, primarily through prompting and supervised fine-tuning. Yet current simulations remain grounded in a behaviorist "demographics in, behavior out" paradigm, focusing on surface-level plausibility. As a result, they often lack internal coherence, causal reasoning, and belief traceability, making them unreliable for modeling how people reason, deliberate, and respond to interventions. To address this, we present a conceptual modeling paradigm, Generative Minds (GenMinds), which draws from cognitive science to support structured belief representations in generative agents. To evaluate such agents, we introduce the RECAP (REconstructing CAusal Paths) framework, a benchmark designed to assess reasoning fidelity via causal traceability, demographic grounding, and intervention consistency. These contributions advance a broader shift: from surface-level mimicry to generative agents that simulate thought, not just language, for social simulations.

Key facts. Cited by 13 papers in the map (data through 14 Oct 2026). Placed in Reasoning for agents, planning and code, one of 17 directions in LLM reasoning. Builds on 39 of its 104 references that are in the map. First posted on arXiv 8 Jun 2025.

Builds on

39 of 104 references.

References · 39

  1. 2022A Generalist Agentcited by 808
  2. 2023Open Problems and Fundamental Limitations of Reinforcement Learning from Human Feedbackcited by 759
  3. 2023Measuring Faithfulness in Chain-of-Thought Reasoningcited by 703
  4. 2023Character-LLM: A Trainable Agent for Role-Playingcited by 339
  5. 2022Selection-Inference: Exploiting Large Language Models for Interpretable Logical Reasoningcited by 334
  6. 2023How Language Model Hallucinations Can Snowballcited by 311
  7. 2023Large Language Models Fail on Trivial Alterations to Theory-of-Mind Taskscited by 292
  8. 2023A Survey of Reinforcement Learning from Human Feedbackcited by 264
  9. 2024From Persona to Personalization: A Survey on Role-Playing Language Agentscited by 212
  10. 2025AgentSociety: Large-Scale Simulation of LLM-Driven Generative Agents Advances Understanding of Human Behaviors and Societycited by 186
  11. 2024A Roadmap to Pluralistic Alignmentcited by 184
  12. 2023Personality Traits in Large Language Modelscited by 169
27 more
  1. 2023From Word Models to World Models: Translating from Natural Language to the Probabilistic Language of Thoughtcited by 119
  2. 2024KG-Agent: An Efficient Autonomous Agent Framework for Complex Reasoning over Knowledge Graphcited by 93
  3. 2020DialoGLUE: A Natural Language Understanding Benchmark for Task-Oriented Dialoguecited by 89
  4. 2024From Individual to Society: A Survey on Social Simulation Driven by Large Language Model-based Agentscited by 82
  5. 2024Extract, Define, Canonicalize: An LLM-based Framework for Knowledge Graph Constructioncited by 70
  6. 2023Towards Better Chain-of-Thought Prompting Strategies: A Surveycited by 62
  7. 2025SocioVerse: A World Model for Social Simulation Powered by LLM Agents and A Pool of 10 Million Real-World Userscited by 55
  8. 2024On the Hardness of Faithful Chain-of-Thought Reasoning in Large Language Modelscited by 43
  9. 2018Stop Explaining Black Box Machine Learning Models for High Stakes Decisions and Use Interpretable Models Insteadcited by 40
  10. 2020Stance Detection Benchmark: How Robust Is Your Stance Detection?cited by 33
  11. 2024Causal Inference with Large Language Model: A Surveycited by 32
  12. 2024Social Science Meets LLMs: How Reliable Are Large Language Models in Social Simulations?cited by 26
  13. 2025On Benchmarking Human-Like Intelligence in Machinescited by 25
  14. 2024LLMs Are Prone to Fallacies in Causal Inferencecited by 23
  15. 2025CounterBench: Evaluating and Improving Counterfactual Reasoning in Large Language Modelscited by 21
  16. 2024Causal Graph Discovery with Retrieval-Augmented Generation based Large Language Modelscited by 16
  17. 2025Language Agents as Digital Representatives in Collective Decision-Makingcited by 16
  18. 2023Language Models with Rationalitycited by 15
  19. 2025MF-LLM: Simulating Population Decision Dynamics via a Mean-Field Large Language Model Frameworkcited by 12
  20. 2025The Hidden Strength of Disagreement: Unraveling the Consensus-Diversity Tradeoff in Adaptive Multi-Agent Systemscited by 11
  21. 2024Large Language Models as Instruments of Power: New Regimes of Autonomous Manipulation and Controlcited by 10
  22. 2024Causal Agent based on Large Language Modelcited by 10
  23. 2025Can Generative AI agents behave like humans? Evidence from laboratory market experimentscited by 10
  24. 2025SPeCtrum: A Grounded Framework for Multidimensional Identity Representation in LLM-Based Agentcited by 8
  25. 2024From Frege to chatGPT: Compositionality in language, cognition, and deep neural networkscited by 7
  26. 2024Counterfactual Causal Inference in Natural Language with Large Language Modelscited by 7
  27. 2024Persona Inconstancy in Multi-Agent LLM Collaboration: Conformity, Confabulation, and Impersonationcited by 5

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 2; 2026: 11.

13 papers.

Sort:

By area:

  1. 2025LLM-Based Social Simulations Require a Boundarycited by 12
  2. 2026Learning User Simulators with Turing Rewardscited by 7
  3. 2026AI Agents Alone Are Not (Yet) Sufficient for Social Simulationcited by 3
  4. 2026Beyond Offline A/B Testing: Context-Aware Agent Simulation for Recommender System Evaluationcited by 3
  5. 2025HugAgent: A Human Simulation Benchmark for Individual-Level Reasoningcited by 2
  6. 2026Persona-Based Simulation of Human Opinion at Population Scalecited by 2
  7. 2026Stable Behavior, Limited Variation: Persona Validity in LLM Agents for Urban Sentiment Perceptioncited by 2
  8. 2026Do LLM-Driven Agents Exhibit Engagement Mechanisms? Controlled Tests of Information Load, Descriptive Norms, and Popularity Cuescited by 0
  9. 2026CogInstrument: Modeling Cognitive Processes for Bidirectional Human-LLM Alignment in Planning Taskscited by 0
  10. 2026In-Situ Behavioral Evaluation for LLM Fairness, Not Standardized-Test Scorescited by 0
  11. 2026ScioMind: Cognitively Grounded Multi-Agent Social Simulation with Anchoring-Based Belief Dynamics and Dynamic Profilescited by 0
  12. 2026A Large Language Model-Driven Agent-Based Modeling Framework with Multi-Round Communication for Simulating Vaccine Opinion Dynamicscited by 0
1 more
  1. 2026An Investigation of the NeurIPS and ICML 2025 Position Trackscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Do Large Language Models Solve the Problems of Agent-Based Modeling? A Critical Review of Generative Social Simulations
  2. 2025Doing Things with Words: Rethinking Theory of Mind Simulation in Large Language Models
  3. 2025Simulating Generative Social Agents via Theory-Informed Workflow Design
  4. 2024Sense and Sensitivity: Evaluating the simulation of social dynamics via Large Language Models
  5. 2025From the Fluency Fallacy to the Micro-to-Macro Validity Gap: Opportunities and Pitfalls of LLMs in Social Simulation
  6. 2023Generative Agents: Interactive Simulacra of Human Behavior
  7. 2025Don't Trust Generative Agents to Mimic Communication on Social Networks Unless You Benchmarked their Empirical Realism
  8. 2023Generative Agent-Based Modeling: Unveiling Social System Dynamics through Coupling Mechanistic Models with Generative Artificial Intelligence