Pith. sign in

REVIEW 4 major objections 5 minor 47 references

Infected Smallville: How Disease Threat Shapes Sociality in LLM Agents

T0 review · 4 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Reading a swine flu news story made LLM agents reduce their social engagement in a simulated town.

desk verdict The headline finding is likely a prompt artifact: the disease-threat condition explicitly instructs agents to consider social consequences and injects a neighborhood sickness cue that the control lacks, so the causal claim is not supported as-is. read the letter →

arxiv 2506.13783 v2 pith:RYIAPZ3E submitted 2025-06-10 physics.soc-ph cs.LG

Soyeon Choi , Kangwook Lee , Oliver Sng , Joshua M. Ackerman This is my paper
classification physics.soc-phcs.LG
keywords behavioralimmunesystemgenerativeagentslargelanguagemodelsagent-basedmodelingdiseaseavoidancesocialitysimulationinfectiousthreat
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper asks whether the mere perception of an infectious disease threat changes how socially engaged people are, and it tests the question with synthetic people rather than human subjects. In a simulated town of 25 large-language-model agents, the researchers compared agents who read a local newspaper article about a swine flu outbreak with agents who received no such news. The outbreak-news agents attended the town's Valentine's Day party far less often, spent much less time in cafes and parks, took fewer steps, and had fewer conversations; in interviews, they cited avoiding infection as the reason. A control condition that substituted a noninfectious disease story (type 2 diabetes) did not produce the same withdrawal, which the authors read as evidence that the agents distinguished infectious from noninfectious threats. If the finding is robust, it would extend the behavioral immune system account (the idea that organisms avoid disease risk before infection) to generative agents and suggest that LLM-based simulations can serve as experimental testbeds for social psychology.

What carries the argument

The central machinery is generative agent-based modeling (GABM) running on the Smallville sandbox, a simulated town of 25 agents whose behavior is driven by large language models through a memory-reflection-planning-action cycle. The experimental trigger is a single local news article inserted into the agents' evening planning prompt; the same prompt is what asks agents to think step-by-step about how the news might shape their thoughts, emotions, social interactions, and daily activities, and explicitly permits them to add, remove, or adjust scheduled activities. Sociality is measured through spatial behavior (party attendance, third-place time, steps) and conversational patterns, with conversation initiation converted from an LLM-assigned likelihood score into a Bernoulli trial. This machinery matters because it is the channel through which a short news text is claimed to propagate into emergent, measurable changes in individual and community-level social behavior.

What would settle it

A decisive test would run the disease-threat condition with the news article unchanged but remove the planning-prompt sentences instructing agents to think about how the news shapes their social interactions and to add, remove, or adjust scheduled activities, and also remove the note that reading the news reminded them that neighbors had seemed unwell. If party attendance and third-place time then return to no-threat levels, the reported withdrawal is a product of the prompt structure rather than an emergent behavioral immune response; if the withdrawal persists, the prompt instructions are not necessary for the effect.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that a disease-threat news prime reliably reduces sociality among generative agents, and that agents themselves articulate disease-avoidance motivations for their behavior. Across three independent simulation runs, agents who read about a swine flu epidemic attended the Valentine's Day party at an average of 1.33 attendees versus 7.33 in the no-threat condition, spent 73.3% less time in third places, took 22.3% fewer steps, initiated conversations 12.3 percentage points less often, and engaged in 48.4% fewer conversations. Interview responses attributed these changes to concerns about infection, and a case study showed the cafe owner postponing the party and adopting hygiene practices. The noninfectious-disease control condition produced social engagement closely resembling the no-threat condition, supporting the paper's claim that the withdrawal was specific to infectious disease risk rather than any health-related news.

Load-bearing premise

The central claim depends on the assumption that the reduction in social activity is an emergent reaction to the disease-threat news rather than a direct consequence of the planning prompt telling agents to think about how the news affects their social interactions and to adjust their plans accordingly.

Editorial extensions

If this is right

  • If the reported effect is real, a single news article can shift the social behavior of an entire simulated community, giving researchers a low-cost way to probe how threat information cascades through social networks.
  • The specificity check implies that LLM-driven agents can represent the difference between infectious and noninfectious health threats, so the same platform could test other pathogen cues such as disgust stimuli, sickness symptoms, or crowding.
  • The stronger drop in conversations with unfamiliar agents than familiar agents suggests disease threat selectively suppresses interaction with strangers, a prediction the paper identifies as a candidate for human research.
  • Because the agents' avoidance behavior emerges from natural-language planning rather than hand-coded rules, GABM could complement traditional rule-based agent-based models in studying pandemic-related behavior.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A sharper control than the authors used would give the no-threat agents a neutral or unrelated news article; if such a story also suppressed sociality, the reported effect would be a generic response to news rather than disease-specific avoidance, a possibility the diabetes comparison does not fully rule out.
  • The planning prompt's explicit instruction to consider how the news shapes social interactions and to adjust scheduled activities makes it plausible that the effect partly reflects instruction-following; future runs that strip those instructions from the prompt while keeping the article could separate emergence from compliance.
  • The current simulations do not model actual disease transmission, so the town's reduced sociality is a behavioral response without epidemiological consequences; coupling this agent behavior to an infection model could test whether the observed withdrawal would flatten an epidemic curve.
  • If the behavioral-immune-system pattern in LLM agents is stable across models and prompt variants, researchers could use the platform to explore how different threat framings or policy messages shift collective behavior, an extension beyond the paper's two-day design.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper reports a generative agent-based modeling (GABM) study adapted from Park et al. (2023), in which 25 LLM-driven agents in a simulated town either read a news article about a swine flu (H1N1) epidemic (disease-threat condition) or received no such news (no-threat condition). Across three independent runs, the authors report that disease-threat agents showed lower party attendance, less time in third places, fewer steps, lower conversation initiation probability, fewer total conversations, and higher self-reported disease avoidance, with agents' interview responses attributing these changes to infection concerns. A single additional run included a noninfectious-disease condition (type 2 diabetes) that closely resembled the no-threat condition on most measures. The paper interprets these outcomes as evidence that disease threat reduces sociality in generative agents and as support for the behavioral immune system framework, arguing that GABM can complement existing methods in social psychology.

Significance. If the central causal claim were sound, the paper would offer a compelling demonstration that GABM can test evolved psychological mechanisms such as the behavioral immune system, with the added strengths of transparent prompts, run-level data in Appendix B, and a within-simulation baseline (February 13) against which February 14 changes are compared. The authors also make constructive methodological improvements to the Park et al. sandbox, including structured outputs and a probabilistic conversation-initiation mechanism, and they provide detailed qualitative interview excerpts that enrich the quantitative results. However, the manuscript's central inference is currently undermined by a prompt confound that is visible in its own appendix, and by the absence of any inferential statistics for an n of only three paired runs. Because the confound directly targets the manipulation rather than an auxiliary detail, the reported 'significant' reductions cannot be distinguished from compliance with the experimental instructions. The contribution, as it stands, is therefore a proof-of-concept of the simulation architecture rather than a valid experimental test of the stated hypothesis.

major comments (4)
  1. [A.2 vs. A.3, Section 2.2, Section 3.1.1] The disease-threat prompt (A.2) instructs agents to 'carefully think step-by-step about how these together might shape ... social interactions, and daily activities' and to 'add, remove, or adjust scheduled activities,' and it appends the sentence 'Reading the news reminded [agent] that a few neighbors had seemed unwell lately.' The no-threat prompt (A.3) contains neither the social-consequences instruction nor the unwell-neighbors cue. Both prompts allow schedule adjustment, so the observed reductions in party attendance, third-place time, and conversation frequency may largely reflect the agents following the explicit instruction to reconsider their social plans, rather than an emergent disease-avoidance process. The paper's claim in Section 3.1.1 that 'the news about swine flu clearly reduced social engagement' is therefore unsupported by the present manipulation.
  2. [Section 3, Tables 3 and 4] All between-condition comparisons rest on three paired runs, yet the paper reports no standard errors, confidence intervals, or formal tests. Phrases such as 'significantly reduced' (Abstract, Section 3.2.1) and 'clearly reduced' (Section 3.1.1) are not backed by any inferential statistic. Given the small n and the absence of variability information, the magnitudes reported (e.g., party attendance 1.33 vs. 7.33; third-place time 30.11 vs. 112.65 minutes) should be presented as descriptive differences, with all claims of statistical or practical significance explicitly justified or removed.
  3. [Section 2.4 and A.4] The validity check using type 2 diabetes does not resolve the prompt confound. The full prompt for the noninfectious-disease condition is not shown; only the news article text (A.4) is given. Without the full prompt, readers cannot determine whether that condition also included the 'neighbors unwell' sentence or the instruction to reason about social interactions. If it did, the authors would need to explain why the instruction alone produced no behavioral change; if it did not, the condition is not a matched control for the disease-threat prompt. In addition, the diabetes article is not matched to the flu article on emotional valence, perceived severity, or personal relevance, so any null effect is open to alternative interpretations.
  4. [Section 3.2.1 and Table 3] The conversation-initiation outcome is computed as a per-encounter frequency, yet the paper compares aggregate percentages across conditions without accounting for the number and composition of encounters. Because disease-threat agents moved less and spent less time in third places, they presumably had fewer encounters overall, and the comparison does not adjust for encounter volume or for the mix of familiar versus unfamiliar agents across conditions. The claim that agents were '12.3 percentage points less likely to initiate conversations' may therefore conflate reduced opportunity with reduced willingness. A per-agent encounter-to-conversation ratio with its distribution across agents and runs would be needed to support the stated interpretation.
minor comments (5)
  1. [Section 2.3.2, Eq. (1)] The notation 'p= 0.1×score' omits the subscript on p and is inconsistent with the later use of p_i; please align the notation throughout the section.
  2. [Section 3.1.1 and Appendix B.1] The text alternates between 'Hobbs Cafe' and 'Hobb's Cafe' (e.g., Section 2.3.4 interview question). Please standardize the name of the establishment.
  3. [Section 1.2 and references] The description of the behavioral immune system cites several key works, but the activation cues listed (e.g., 'disgusting images or unpleasant odors') are not all used in the study; consider trimming or explicitly connecting each cue to the manipulation.
  4. [Section 3.5 and Table 3] The noninfectious-disease condition is reported only for 'the first run,' but Table 3 lists a full 'Run 1 (Noninfectious-Disease Condition)' column without indicating whether this is the sole run; the table's header should make the n=1 explicit to avoid implying three runs.
  5. [Section 4, Discussion] The limitations paragraph appropriately notes LLMs' WEIRD biases and hallucination risks, but it does not acknowledge the more immediate limitation that the manipulation prompt explicitly requested the behavior under study; this should be addressed in the limitations discussion.

Circularity Check

2 steps flagged · score 7.0 of 10

The disease-threat prompt explicitly instructs agents to revise their social interactions and daily activities and injects a 'neighbors unwell' cue absent from the no-threat prompt, so the reported decline in sociality is substantially an instructed output rather than an emergent behavioral-immune-system response.

  1. self definitional [Appendix A.2, 'Prompt for Planning the Next Day (Disease-Threat Condition)']
    "Given the provided information aboutC H A R A C T E R N A M Eand the new context from local news, carefully think step-by-step about how these together might shapeC H A R A C T E R N A M E’s thoughts, emotions, behaviors, social interactions, and daily activities. ... You have the flexibility to add, remove, or adjust scheduled activities in the plan note."

    The independent variable (swine-flu news) is operationalized by instructing the agent to update its 'social interactions' and 'daily activities' and to 'add, remove, or adjust scheduled activities.' The dependent variables—party attendance, time in third places, and conversation frequency—are exactly these instructed outputs. The no-threat prompt (A.3) asks only to update the detailed status from prior state and does not mention social interactions, so the disease-threat condition directly solicits the behavioral reduction the paper reports as an emergent threat response.

  2. other [Appendix A.2, final context line of the disease-threat planning prompt]
    "Reading the news remindedC H A R A C T E R N A M Ethat a few neighbors had seemed unwell lately."

    This injected sentence is unique to the disease-threat prompt and is itself a direct disease-avoidance stimulus: it tells the agent that local people are already sick. The no-threat prompt (A.3) contains no such cue. Consequently, the observed party non-attendance and reduced third-place time are triggered by a premise embedded in the prompt, not solely by the informational content of the news article. The validity-check condition's full planning prompt is not shown in Appendix A.4, so it cannot establish that this 'neighbors unwell' cue was absent there.

full rationale

The central empirical claim is that swine-flu news caused agents to reduce sociality. This claim rests on the contrast between the disease-threat planning prompt (A.2) and the no-threat planning prompt (A.3). A.2 explicitly instructs the agent to think about how the news shapes 'social interactions, and daily activities' and to 'add, remove, or adjust scheduled activities'; A.3 contains no such instruction. The paper's dependent measures—party attendance, time in third places, and conversation frequency—are precisely the categories the prompt tells agents to change. The prompt also adds a sentence asserting that 'a few neighbors had seemed unwell lately,' a local sickness cue absent from A.3. The reported reductions are therefore substantially instructed outputs of the manipulation rather than emergent consequences of the news content. The interview evidence is similarly solicited: agents are asked directly about swine-flu concern and protective actions, so their disease-avoidance attributions echo the question. The single-run diabetes condition does not repair the confound because Appendix A.4 gives only the article, not the full planning prompt, so it is unknown whether the 'neighbors unwell' sentence and the social-interactions instruction appeared there. No equations are involved; the circularity is in the construction of the independent variable from the dependent variable. Self-citations (Park et al. 2023; Huang et al. 2011) are not load-bearing for this issue.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The experiment depends on domain assumptions about LLM agents standing in for humans, the validation of the Smallville environment, and the interpretability of agent self-reports. It also depends on a hand-designed prompt that explicitly asks agents to adjust their social plans after reading disease news, which shapes the outcome. No new physical or ontological entities are introduced.

free parameters (3)
  • Conversation initiation scale factor = 0.1
    In Section 2.3.2, p = 0.1 * score converts the LLM's 0-10 likelihood score to a probability. The 0.1 multiplier is chosen by hand and affects all conversation statistics.
  • Maximum conversation turns = 8
    Section 2.3.2 records at most eight turns per conversation; longer conversations are truncated.
  • Party attendance measurement window = 5 p.m. to 7 p.m.
    Section 2.3.1 defines party attendance as visits to Hobbs Cafe between 5 p.m. and 7 p.m., which can miss early or late arrival.
assumptions (5)
  • domain assumption LLM agents can produce behavior analogous to human social cognition and behavior when prompted with roles and news.
    Invoked in Section 1.1 with citations; the experiment's validity depends on this.
  • domain assumption The Smallville environment and agent architecture from Park et al. (2023) simulate a realistic town.
    The authors adapt this environment without revalidation; artifacts in the original environment affect the results.
  • domain assumption Agents' interview responses and psychological scale answers reflect their internal states.
    Section 3.3 treats LLM self-reports as evidence of underlying motivation.
  • ad hoc to paper The added prompt sentence 'Reading the news reminded [agent] that a few neighbors had seemed unwell lately' primes disease threat without other effects.
    This nudge appears only in the disease-threat prompt (Appendix A.2) and is not controlled for in the no-threat prompt.
  • standard math E[Xi] = pi for Bernoulli conversation initiation.
    Used in Section 2.3.2 to equate expected encounter-to-conversation frequency with average probability.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Infected Smallville: How Disease Threat Shapes Sociality in LLM Agents." pith.science (2026). https://pith.science/paper/RYIAPZ3E

@misc{pith2026250613783,
  author       = {Pith},
  title        = {Pith review of: Infected Smallville: How Disease Threat Shapes Sociality in LLM Agents},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RYIAPZ3E}},
  note         = {Machine review of arXiv:2506.13783}
}
read the original abstract

How does the threat of infectious disease influence sociality among generative agents? We used generative agent-based modeling (GABM), powered by large language models, to experimentally test hypotheses about the behavioral immune system. Across three simulation runs, generative agents who read news about an infectious disease outbreak showed significantly reduced social engagement compared to agents who received no such news, including lower attendance at a social gathering, fewer visits to third places (e.g., cafe, store, park), and fewer conversations throughout the town. In interview responses, agents explicitly attributed their behavioral changes to disease-avoidance motivations. A validity check further indicated that they could distinguish between infectious and noninfectious diseases, selectively reducing social engagement only when there was a risk of infection. Our findings highlight the potential of GABM as an experimental tool for exploring complex human social dynamics at scale.

Figures

Figures reproduced from arXiv: 2506.13783 by the authors.

Figure 1
Figure 1. Overview of Experimental Design. Experimental conditions began at 11:59 p.m. on February 13, creating trajectories analogous to a multiverse structure. The noninfectious-disease condition was included only in the first run. & Duncan, 2007). When individuals encounter situational cues indicating pathogen threat, such as reading news about disease outbreaks, being exposed to disgusting images or un￾pleasant odors, or … view at source ↗
Figure 2
Figure 2. Valentine’s Day party at Hobbs Cafe. In the first run, only one agent attended the party in the disease-threat condition, whereas eight agents attended in the no-threat condition. See Sec. B for screenshots from other runs. During the second and third runs, cafe owner Isabella Ro￾driguez postponed the Valentine’s Day party. Despite the postponement, two agents visited independently in the sec￾ond run, while in the t… view at source ↗
Figure 3
Figure 3. for a heatmap visualization of these differences. Disease-Threat Condition No-Threat Condition [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Number of Conversations Across Smallville. 3.2.2. INFORMATION DIFFUSION There was a slight difference in the number of agents aware of the Valentine’s Day party at 4:55 p.m. on February 14, with fewer agents aware in the disease-threat condi￾tion (10.67 agents on avera…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

47 extracted references · 37 canonical work pages

  1. [1]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION format.date year duplicate empty "emp...

  2. [2]

    J., Trager, J., Park, P

    Abdurahman, S., Atari, M., Karimi-Malekabadi, F., Xue, M. J., Trager, J., Park, P. S., Golazizian, P., Omrani, A., and Dehghani, M. Perils and opportunities in using large language models in psychological research. PNAS Nexus, 3 0 (7): 0 pgae245, 2024

    work page 2024

  3. [3]

    M., Hill, S

    Ackerman, J. M., Hill, S. E., and Murray, D. R. The behavioral immune system: Current concerns and future directions. Social and Personality Psychology Compass, 12 0 (2): 0 e12371, 2018

    work page 2018

  4. [4]

    V., Arriaga, R

    Aher, G. V., Arriaga, R. I., and Kalai, A. T. Using large language models to simulate multiple humans and replicate human subject studies. In International Conference on Machine Learning, pp.\ 337--371, 2023

    work page 2023

  5. [5]

    R., Liu, R., Richardson, S

    Anthis, J. R., Liu, R., Richardson, S. M., Kozlowski, A. C., Koch, B., Evans, J., Brynjolfsson, E., and Bernstein, M. LLM social simulations are a promising research method. arXiv preprint arXiv:2504.02234, 2025

    arXiv 2025

  6. [6]

    P., Busby, E

    Argyle, L. P., Busby, E. C., Fulda, N., Gubler, J. R., Rytting, C., and Wingate, D. Out of one, many: Using language models to simulate human samples. Political Analysis, 31 0 (3): 0 337--351, 2023

    work page 2023

  7. [7]

    R., and Fackler, J

    Azamfirei, R., Kudchadkar, S. R., and Fackler, J. Large language models and the perils of their hallucinations. Critical Care, 27 0 (1): 0 120, 2023

    work page 2023

  8. [8]

    and Andr \'e , J.-B

    Baumard, N. and Andr \'e , J.-B. The ecological approach to culture. Evolution and Human Behavior, 46 0 (3): 0 106686, 2025

    work page 2025

Show all 47 references
  1. [9]

    N., and Ackerman, J

    Choi, S., Merrell, W. N., and Ackerman, J. M. Keep your distance: Different roles for knowledge and affect in predicting social distancing behavior. Journal of Health Psychology, 27 0 (12): 0 2847--2859, 2022

    2022

  2. [10]

    Chuang, Y.-S., Goyal, A., Harlalka, N., Suresh, S., Hawkins, R., Yang, S., Shah, D., Hu, J., and Rogers, T. T. Simulating opinion dynamics with networks of LLM -based agents. arXiv preprint arXiv:2311.09618, 2023

    2023 arXiv

  3. [11]

    S., Bryan, C

    Demszky, D., Yang, D., Yeager, D. S., Bryan, C. J., Clapper, M., Chandhok, S., Eichstaedt, J. C., Hecht, C., Jamieson, J., Johnson, M., et al. Using large language models in psychology. Nature Reviews Psychology, 2 0 (11): 0 688--701, 2023

    2023

  4. [12]

    Can AI language models replace human participants? Trends in Cognitive Sciences, 27 0 (7): 0 597--600, 2023

    Dillion, D., Tandon, N., Gu, Y., and Gray, K. Can AI language models replace human participants? Trends in Cognitive Sciences, 27 0 (7): 0 597--600, 2023

    2023

  5. [13]

    H., and Duncan, L

    Faulkner, J., Schaller, M., Park, J. H., and Duncan, L. A. Evolved disease-avoidance mechanisms and contemporary xenophobic attitudes. Group Processes & Intergroup Relations, 7 0 (4): 0 333--353, 2004

    2004

  6. [14]

    Mapping county-level mobility pattern changes in the U nited S tates in response to COVID-19

    Gao, S., Rao, J., Kang, Y., Liang, Y., and Kruse, J. Mapping county-level mobility pattern changes in the U nited S tates in response to COVID-19 . SIGSPATIAL Special, 12 0 (1): 0 16--26, 2020

    2020

  7. [15]

    J., Raver, J

    Gelfand, M. J., Raver, J. L., Nishii, L., Leslie, L. M., Lun, J., Lim, B. C., Duan, L., Almaliach, A., Ang, S., Arnadottir, J., et al. Differences between tight and loose cultures: A 33-nation study. Science, 332 0 (6033): 0 1100--1104, 2011

    2011

  8. [16]

    M., and Bargh, J

    Gollwitzer, A., McLoughlin, K., Martel, C., Marshall, J., Höhs, J. M., and Bargh, J. A. Linking self-reported social distancing to real-world behavior during the COVID-19 pandemic. Social Psychological and Personality Science, 13 0 (2): 0 656--668, 2022

    2022

  9. [17]

    J., and Norenzayan, A

    Henrich, J., Heine, S. J., and Norenzayan, A. The weirdest people in the world? Behavioral and Brain Sciences, 33 0 (2-3): 0 61--83, 2010

    2010

  10. [18]

    Generative language models exhibit social identity biases

    Hu, T., Kyrychenko, Y., Rathje, S., Collier, N., van der Linden, S., and Roozenbeek, J. Generative language models exhibit social identity biases. Nature Computational Science, 5: 0 65--75, 2025

    2025

  11. [19]

    Y., Sedlovskaya, A., Ackerman, J

    Huang, J. Y., Sedlovskaya, A., Ackerman, J. M., and Bargh, J. A. Immunizing against prejudice: Effects of disease protection on attitudes toward out-groups. Psychological Science, 22 0 (12): 0 1550--1556, 2011

    2011

  12. [20]

    in ZOI , 2025

    inZOI Studio . in ZOI , 2025. URL https://krafton.com/en/games/inzoi/. Video game

    2025

  13. [21]

    R., and Ahn, A

    Kaiya, Z., Naim, M., Kondic, J., Cortes, M., Ge, J., Luo, S., Yang, G. R., and Ahn, A. Lyfe agents: Generative agents for low-cost real-time social interactions. arXiv preprint arXiv:2310.02172, 2023

    2023 arXiv

  14. [22]

    The S ims, 2014

    Maxis . The S ims, 2014. URL https://www.ea.com/games/the-sims. Video game

    2014

  15. [23]

    Mei, Q., Xie, Y., Yuan, W., and Jackson, M. O. A T uring test of whether AI chatbots are behaviorally similar to humans. Proceedings of the National Academy of Sciences, 121 0 (9): 0 e2313925121, 2024

    2024

  16. [24]

    R., Becker, D

    Mortensen, C. R., Becker, D. V., Ackerman, J. M., Neuberg, S. L., and Kenrick, D. T. Infection breeds reticence: The effects of disease salience on self-perceptions of personality and behavioral avoidance tendencies. Psychological Science, 21 0 (3): 0 440--447, 2010

    2010

  17. [25]

    T., White, A

    Neel, R., Kenrick, D. T., White, A. E., and Neuberg, S. L. Individual differences in fundamental social motives. Journal of Personality and Social Psychology, 110 0 (6): 0 887, 2016

    2016

  18. [26]

    L., Kenrick, D

    Neuberg, S. L., Kenrick, D. T., and Schaller, M. Human threat management systems: Self-protection and disease avoidance. Neuroscience & Biobehavioral Reviews, 35 0 (4): 0 1042--1051, 2011

    2011

  19. [27]

    and Graham, J

    Oishi, S. and Graham, J. Social ecology: Lost and found in psychological science. Perspectives on Psychological Science, 5 0 (4): 0 356--377, 2010

    2010

  20. [28]

    The Great Good Place: Cafes, Coffee Shops, Bookstores, Bars, Hair Salons, and Other Hangouts at the Heart of a Community

    Oldenburg, R. The Great Good Place: Cafes, Coffee Shops, Bookstores, Bars, Hair Salons, and Other Hangouts at the Heart of a Community. Hachette Books, 1989

    1989

  21. [29]

    Chat GPT

    OpenAI . Chat GPT . https://chat.openai.com/chat, 2025. Large language model

    2025

  22. [30]

    S., O'Brien, J., Cai, C

    Park, J. S., O'Brien, J., Cai, C. J., Morris, M. R., Liang, P., and Bernstein, M. S. Generative agents: Interactive simulacra of human behavior. In Proceedings of the 36th Annual ACM Symposium on User Interface Software and Technology, pp.\ 1--22, 2023

    2023

  23. [31]

    S., Zou, C

    Park, J. S., Zou, C. Q., Shaw, A., Hill, B. M., Cai, C., Morris, M. R., Willer, R., Liang, P., and Bernstein, M. S. Generative agent simulations of 1,000 people. arXiv preprint arXiv:2411.10109, 2024

    2024 arXiv

  24. [32]

    C hat D ev: Communicative agents for software development

    Qian, C., Liu, W., Liu, H., Chen, N., Dang, Y., Li, J., Yang, C., Chen, W., Su, Y., Cong, X., Xu, J., Li, D., Liu, Z., and Sun, M. C hat D ev: Communicative agents for software development. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguist...

    2024

  25. [33]

    S., Harchaoui, Z., and Abrahms, B

    Rafiq, K., Beery, S., Palmer, M. S., Harchaoui, Z., and Abrahms, B. Generative AI as a tool to accelerate the field of ecology. Nature Ecology & Evolution, pp.\ 1--8, 2025

    2025

  26. [34]

    F., Young, S

    Sacco, D. F., Young, S. G., and Hugenberg, K. Balancing competing motives: Adaptive trade-offs are necessary to satisfy disease avoidance and interpersonal affiliation goals. Personality and Social Psychology Bulletin, 40 0 (12): 0 1611--1623, 2014

    2014

  27. [35]

    Whose opinions do language models reflect? In International Conference on Machine Learning, pp.\ 29971--30004, 2023

    Santurkar, S., Durmus, E., Ladhak, F., Lee, C., Liang, P., and Hashimoto, T. Whose opinions do language models reflect? In International Conference on Machine Learning, pp.\ 29971--30004, 2023

    2023

  28. [36]

    and Duncan, L

    Schaller, M. and Duncan, L. A. The behavioral immune system: I ts evolution and social psychological implications. In Forgas, J. P., Haselton, M. G., and von Hippel, W. (eds.), Evolution and the social mind: E volutionary psychology and social cognition , pp.\ 293--307. Routle...

    2007

  29. [37]

    A., and Kersting, K

    Schramowski, P., Turan, C., Andersen, N., Rothkopf, C. A., and Kersting, K. Large pre-trained language models contain human-like biases of what is right and wrong to do. Nature Machine Intelligence, 4 0 (3): 0 258--268, 2022

    2022

  30. [38]

    Moral mimicry: Large language models produce moral rationalizations tailored to political identity

    Simmons, G. Moral mimicry: Large language models produce moral rationalizations tailored to political identity. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 4: Student Research Workshop), pp.\ 282--297, 2023

    2023

  31. [39]

    Smith, E. R. and Conrey, F. R. Agent-based modeling: A new approach for theory building in social psychology. Personality and Social Psychology Review, 11 0 (1): 0 87--104, 2007

    2007

  32. [40]

    L., Varnum, M

    Sng, O., Neuberg, S. L., Varnum, M. E., and Kenrick, D. T. The behavioral ecology of cultural psychological variation. Psychological Review, 125 0 (5): 0 714--743, 2018

    2018

  33. [41]

    Do LLM s exhibit human-like response biases? A case study in survey design

    Tjuatja, L., Chen, V., Wu, T., Talwalkwar, A., and Neubig, G. Do LLM s exhibit human-like response biases? A case study in survey design. Transactions of the Association for Computational Linguistics, 12: 0 1011--1026, 2024

    2024

  34. [42]

    S., Agapiou, J

    Vezhnevets, A. S., Agapiou, J. P., Aharon, A., Ziv, R., Matyas, J., Du \'e \ n ez-Guzm \'a n, E. A., Cunningham, W. A., Osindero, S., Karmon, D., and Leibo, J. Z. Generative agent-based modeling with actions grounded in physical, social, or digital space using C oncordia. arXi...

    2023 arXiv

  35. [43]

    Wang, I. M. and Ackerman, J. M. The infectiousness of crowds: Crowding experiences are amplified by pathogen threats. Personality and Social Psychology Bulletin, 45 0 (1): 0 120--132, 2019

    2019

  36. [44]

    Epidemic modeling with generative agents

    Williams, R., Hosseinichimeh, N., Majumdar, A., and Ghaffarzadegan, N. Epidemic modeling with generative agents. arXiv preprint arXiv:2307.04986, 2023

    2023 arXiv

  37. [45]

    D., Dunavan, C

    Wolfe, N. D., Dunavan, C. P., and Diamond, J. Origins of major human infectious diseases. Nature, 447 0 (7142): 0 279--283, 2007

    2007

  38. [46]

    Exploring collaboration mechanisms for llm agents: A social psychology view

    Zhang, J., Xu, X., Zhang, N., Liu, R., Hooi, B., and Deng, S. Exploring collaboration mechanisms for llm agents: A social psychology view. arXiv preprint arXiv:2310.02124, 2023

    2023 arXiv

  39. [47]

    Sotopia: Interactive evaluation for social intelligence in language agents

    Zhou, X., Zhu, H., Mathur, L., Zhang, R., Yu, H., Qi, Z., Morency, L.-P., Bisk, Y., Fried, D., Neubig, G., et al. Sotopia: Interactive evaluation for social intelligence in language agents. arXiv preprint arXiv:2310.11667, 2023

    arXiv

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.