mlsanityData to 7 Oct 2026

Improving Behavioral Alignment in LLM Social Simulations via Context Formation and Navigation

arXiv 2601.01546 ↗First posted 4 Jan 2026cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Large language models (LLMs) are increasingly used to simulate human behavior in experimental settings, but they systematically diverge from human decisions in complex decision-making environments, where participants must anticipate others' actions and form beliefs based on observed behavior. We propose a two-stage framework for improving behavioral alignment. The first stage, context formation, explicitly specifies the experimental design to establish an accurate representation of the decision task and its context. The second stage, context navigation, guides the reasoning process within that representation to make decisions. We validate this framework through a focal replication of a sequential purchasing game with quality signaling (Kremer and Debo, 2016), extending to a crowdfunding game with costly signaling (Cason et al., 2025) and a demand-estimation task (Gui and Toubia, 2025) to test generalizability across decision environments. Across four state-of-the-art (SOTA) models (GPT-4o, GPT-5, Claude-4.0-Sonnet-Thinking, DeepSeek-R1), we find that complex decision-making environments require both stages to achieve behavioral alignment with human benchmarks, whereas the simpler demand-estimation task requires only context formation. Our findings clarify when each stage is necessary and provide a systematic approach for designing and diagnosing LLM social simulations as complements to human subjects in behavioral research.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 26 of its 82 references that are in the map. First posted on arXiv 4 Jan 2026.

Builds on

26 of 82 references.

References · 26

  1. 2022Training language models to follow instructions with human feedbackcited by 15,712
  2. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,355
  3. 2022Large Language Models are Zero-Shot Reasonerscited by 4,323
  4. 2017Deep reinforcement learning from human preferencescited by 3,846
  5. 2021Finetuned Language Models Are Zero-Shot Learnerscited by 3,747
  6. 2019Fine-Tuning Language Models from Human Preferencescited by 2,124
  7. 2024Tulu 3: Pushing Frontiers in Open Language Model Post-Trainingcited by 988
  8. 2022Out of One, Many: Using Language Models to Simulate Human Samplescited by 737
  9. 2022Using Large Language Models to Simulate Multiple Humans and Replicate Human Subject Studiescited by 506
  10. 2022Social Simulacra: Creating Populated Prototypes for Social Computing Systemscited by 300
  11. 2023The Impact of AI on Developer Productivity: Evidence from GitHub Copilotcited by 264
  12. 2023Playing repeated games with Large Language Modelscited by 213
14 more
  1. 2023Generative artificial intelligence enhances creativity but reduces the diversity of novel contentcited by 165
  2. 2023Simulating Opinion Dynamics with Networks of LLM-based Agentscited by 153
  3. 2024OASIS: Open Agent Social Interaction Simulations with One Million Agentscited by 127
  4. 2025LLM Social Simulations Are a Promising Research Methodcited by 97
  5. 2023The Challenge of Using LLMs to Simulate Human Behavior: A Causal Inference Perspectivecited by 57
  6. 2023Do LLM Agents Exhibit Social Behavior?cited by 53
  7. 2023AI-Augmented Surveys: Leveraging Large Language Models and Surveys for Opinion Predictioncited by 46
  8. 2025TwinMarket: A Scalable Behavioral and Social Simulation for Financial Marketscited by 46
  9. 2023Automated Paper Screening for Clinical Reviews Using Large Language Modelscited by 29
  10. 2024Take Caution in Using LLMs as Human Surrogates: Scylla Ex Machinacited by 29
  11. 2025AgentA/B: Automated and Scalable Web A/BTesting with Interactive LLM Agentscited by 19
  12. 2024LOLA: LLM-Assisted Online Learning Algorithm for Content Experimentscited by 18
  13. 2024Large Language Models for Market Research: A Data-augmentation Approachcited by 18
  14. 2024The Value of AI-Generated Metadata for UGC Platforms: Evidence from a Large-scale Field Experimentcited by 5

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Interpreting and Steering LLM Agents for Social Simulations
  2. 2025SimBench: Benchmarking the Ability of Large Language Models to Simulate Human Behaviors
  3. 2025Noise, Adaptation, and Strategy: Assessing LLM Fidelity in Decision-Making
  4. 2024Reproducing and Extending Experiments in Behavioral Strategy with Large Language Models
  5. 2024Are LLMs Socially Adaptive? Contrasting Belief Evolution in Large Language Models and Humans
  6. 2025GTAlign: Game-Theoretic Alignment of LLM Assistants for Social Welfare
  7. 2026Should LLM Agents Decide in Social Simulations? Comparing Finite-State and LLM-Based Decision Policies
  8. 2025Computational Basis of LLM's Decision Making in Social Simulation