pith. sign in

arxiv: 2605.15812 · v1 · pith:6SN5TYPLnew · submitted 2026-05-15 · 💻 cs.HC · cs.AI

Toward Natural and Companionable Virtual Agents via Cross-Temporal Emotional Modeling

Pith reviewed 2026-05-20 19:01 UTC · model grok-4.3

classification 💻 cs.HC cs.AI
keywords cross-temporal emotion modelingvirtual companion agentsemotional state modelinglong-term interactionsconversational agentsnaturalnesscoherenceemotional harmony
0
0 comments X

The pith

Cross-Temporal Emotion Modeling creates a closed loop that lets virtual agents evolve emotional states from past interactions and feedback to improve long-term naturalness.

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

Current conversational agents often feel episodic and inauthentic because their behaviors do not influence an internal emotional state and emotions do not shape future actions. Cross-Temporal Emotion Modeling addresses this by linking long-term behavioral history to immediate emotional expressions through a continuous cycle of updates from experiences and user input. If effective, this approach would allow agents to reflect on shared history and anticipate user responses, fostering more authentic companionship over extended periods. The authors test this by building Auri, an agent on a messaging platform, and observing better ratings for naturalness and coherence after three weeks of use.

Core claim

The paper presents Cross-Temporal Emotion Modeling (CTEM) as a framework that establishes a closed loop: past experiences update an evolving emotional state, this state conditions immediate interactions, and user feedback revises both memory and emotional state to enable reflection and anticipation. When implemented in the Auri agent, this modeling leads to measurable improvements in perceived naturalness, coherence, and emotional harmony during a 21-day in-the-wild study.

What carries the argument

Cross-Temporal Emotion Modeling (CTEM), a framework that connects an agent's long-term memory and evolving emotional state to shape each interaction while incorporating user feedback to update both.

Load-bearing premise

The 21-day in-the-wild study with Auri on an instant-messaging platform provides a valid measure of long-term companion-like naturalness and emotional harmony without major confounds such as novelty effects, self-selection bias, or lack of blinded controls.

What would settle it

A follow-up study in which users rate emotional harmony and coherence the same for CTEM agents and standard agents after the first week of use would show the framework does not produce sustained gains.

Figures

Figures reproduced from arXiv: 2605.15812 by Feier Qin, Haibin Huang, Hanyao Wang, Xiao Li, Xiaoyu Wang, Yan Lu, Yi Zheng, Yuan Zhang.

Figure 1
Figure 1. Figure 1: We present Auri, a lightweight companion agent designed to foster long-term emotional connections through cross-temporal emotional modeling. Auri is able to deliver contextually coherent and emotionally resonant interactions over time. Abstract Recent advances in foundation models have enabled conversational agents that aim for sustained companionship rather than mere task completion. Yet most still remain… view at source ↗
Figure 2
Figure 2. Figure 2: System overview of Auri, showing the CTEM framework with front-end and back-end components. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Examples of cross-temporal interaction under CTEM across two scenarios. BGI refers to Behavior [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Mean ratings of trait contributions to perceived [PITH_FULL_IMAGE:figures/full_fig_p009_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Distribution of usage contexts across groups in [PITH_FULL_IMAGE:figures/full_fig_p009_5.png] view at source ↗
read the original abstract

Recent advances in foundation models have enabled conversational agents that aim for sustained companionship rather than mere task completion. Yet most still remain unable to support natural, long-term companion-like interactions, resulting in experiences that feel episodic and inauthentic. We argue that current agents overlooked cross-temporal modeling of agents' social behaviors and internal emotions: generated behaviors rarely influence an agent's emotional state, and emotional states seldom shape subsequent behaviors. We present Cross-Temporal Emotion Modeling (CTEM), a framework that links long-term behavioral history to moment-to-moment emotional expression. CTEM establishes a closed loop where past experiences update an evolving emotional state; this state conditions immediate interactions; and user feedback continually revises both memory and emotional state, enabling reflection and anticipation. We instantiate CTEM as Auri, a companion agent on an instant-messaging platform, and report a 21-day in-the-wild study showing that CTEM shows improvements in perceived naturalness, coherence, and emotional harmony.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript proposes Cross-Temporal Emotion Modeling (CTEM), a framework that links long-term behavioral history to moment-to-moment emotional expression in virtual agents via a closed loop: past experiences update an evolving emotional state, which conditions immediate interactions, and user feedback revises both memory and emotional state to enable reflection and anticipation. CTEM is instantiated as Auri, a companion agent on an instant-messaging platform, and evaluated in a 21-day in-the-wild study that reports improvements in perceived naturalness, coherence, and emotional harmony.

Significance. If the empirical claims hold after addressing validation gaps, CTEM could meaningfully advance HCI research on sustained companionship in conversational agents by explicitly modeling cross-temporal emotional dynamics, a dimension often missing in current foundation-model systems. The closed-loop structure provides a clear conceptual contribution that could guide future agent architectures.

major comments (2)
  1. [Abstract] Abstract: the claim that the 21-day study 'shows improvements' in naturalness, coherence, and emotional harmony supplies no participant numbers, metrics, statistical tests, control conditions, or exclusion criteria, so it is impossible to verify whether the gains are supported by the data or attributable to the CTEM closed loop.
  2. [Evaluation] User study description: without a control condition, pre/post baseline, or statistical isolation of the emotional-modeling component, improvements cannot be confidently attributed to CTEM rather than confounds such as novelty effects or self-selection bias; this directly undermines the central claim that the cross-temporal loop produces the observed outcomes.
minor comments (2)
  1. [Framework] Provide pseudocode or a diagram for the exact state-update and feedback-revision rules in CTEM to make the framework reproducible.
  2. [Evaluation] Clarify how 'emotional harmony' was operationalized in the study questionnaire and whether inter-rater reliability was assessed.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive and detailed feedback. The comments highlight important issues regarding the presentation of our empirical results and the strength of causal claims. We address each point below and indicate the revisions we will make to the manuscript.

read point-by-point responses
  1. Referee: [Abstract] Abstract: the claim that the 21-day study 'shows improvements' in naturalness, coherence, and emotional harmony supplies no participant numbers, metrics, statistical tests, control conditions, or exclusion criteria, so it is impossible to verify whether the gains are supported by the data or attributable to the CTEM closed loop.

    Authors: We agree that the abstract is too concise and does not supply the quantitative details needed for immediate verification. The Evaluation section of the manuscript reports the full study protocol, including participant numbers, the specific metrics and scales employed, the statistical tests performed, and exclusion criteria. To resolve this, we will revise the abstract to concisely include these elements (participant count, key metrics, statistical outcomes) while preserving its length constraints. This change will allow readers to assess the reported improvements without first consulting the full text. revision: yes

  2. Referee: [Evaluation] User study description: without a control condition, pre/post baseline, or statistical isolation of the emotional-modeling component, improvements cannot be confidently attributed to CTEM rather than confounds such as novelty effects or self-selection bias; this directly undermines the central claim that the cross-temporal loop produces the observed outcomes.

    Authors: We acknowledge that the absence of a control condition limits strong causal attribution to the CTEM closed loop specifically, and that factors such as novelty or self-selection could contribute to the observed changes. The study was intentionally designed as a 21-day in-the-wild deployment to examine sustained, real-world companion interactions rather than a short-term lab experiment. Pre- and post-study questionnaires were administered to capture longitudinal shifts. In revision we will (1) add an explicit Limitations subsection that discusses potential confounds and the exploratory nature of the design, (2) moderate the language in the abstract and conclusion from “shows improvements” to “indicates improvements consistent with” the CTEM framework, and (3) outline directions for future controlled studies that could isolate the emotional-modeling component. These revisions address the concern without altering the core contribution of the closed-loop architecture. revision: partial

Circularity Check

0 steps flagged

No circularity: CTEM is a conceptual framework with independent empirical evaluation

full rationale

The paper presents CTEM as a modeling framework that establishes a closed loop between past experiences, evolving emotional state, immediate interactions, and user feedback. This is instantiated as Auri and supported by a 21-day in-the-wild study reporting improvements in naturalness, coherence, and emotional harmony. No equations, fitted parameters renamed as predictions, self-citations, or uniqueness theorems are described that would reduce the framework or results to inputs by construction. The derivation chain is self-contained as a proposal plus separate user study, with no load-bearing steps that collapse into self-definition or fitted data.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 1 invented entities

Based solely on the abstract, the central claim rests on the introduction of an evolving emotional state as a modeling construct and the assumption that a closed behavioral-emotional-feedback loop produces measurable improvements; no explicit free parameters or prior axioms are stated.

axioms (1)
  • domain assumption Past behaviors and user feedback can be used to maintain and update an internal emotional state that then shapes future agent responses
    This is the core premise of the CTEM framework described in the abstract.
invented entities (1)
  • evolving emotional state no independent evidence
    purpose: To link long-term behavioral history to moment-to-moment emotional expression and enable reflection and anticipation
    New modeling construct introduced by CTEM; no independent evidence or falsifiable prediction outside the framework is mentioned in the abstract.

pith-pipeline@v0.9.0 · 5712 in / 1519 out tokens · 61781 ms · 2026-05-20T19:01:30.157927+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

78 extracted references · 78 canonical work pages · 3 internal anchors

  1. [1]

    David Adam. 2025. Supportive? Addictive? Abusive? How AI companions affect our mental health.Nature641, 8062 (2025), 296–298

  2. [2]

    Sadia Azmin Anisha, Arkendu Sen, and Chris Bain. 2024. Evaluating the potential and pitfalls of AI-powered conversational agents as humanlike virtual health carers in the remote management of noncommunicable diseases: scoping review. Journal of medical Internet research26 (2024), e56114

  3. [3]

    1987.Unraveling the Mystery of Health: How People Manage Stress and Stay Well

    Aaron Antonovsky. 1987.Unraveling the Mystery of Health: How People Manage Stress and Stay Well. Jossey-Bass, San Francisco, CA

  4. [4]

    Roy F Baumeister, Kathleen D Vohs, C Nathan DeWall, and Liqing Zhang. 2007. The psychology of emotion regulation: An integrative review.Review of General Psychology11, 2 (2007), 179–199

  5. [5]

    Maik Beege and Sascha Schneider. 2023. Emotional design of pedagogical agents: the influence of enthusiasm and model-observer similarity.Educational technol- ogy research and development71, 3 (2023), 859–880

  6. [6]

    Timothy Bickmore, Daniel Schulman, and Langxuan Yin. 2010. Maintaining engagement in long-term interventions with relational agents.Applied Artificial Intelligence24, 6 (2010), 648–666

  7. [7]

    Timothy W Bickmore and Tonio Giorgino. 2006. Health dialog systems for patients and consumers.Journal of Biomedical Informatics39, 5 (2006), 556–571

  8. [8]

    Bickmore and Rosalind W

    Timothy W. Bickmore and Rosalind W. Picard. 2005. Establishing and Maintaining Long-Term Human–Computer Relationships.ACM Transactions on Computer– Human Interaction (TOCHI)12, 2 (2005), 293–327. doi:10.1145/1067860.1067867

  9. [9]

    Sylvie Borau, Tobias Otterbring, Sandra Laporte, and Samuel Fosso Wamba. 2021. The most human bot: Female gendering increases humanness perceptions of bots and acceptance of AI.Psychology & Marketing38, 7 (2021), 1052–1068

  10. [10]

    Petter Bae Brandtzaeg, Marita Skjuve, and Asbjørn Følstad. 2022. My AI friend: How users of a social chatbot understand their human–AI friendship.Human Communication Research48, 3 (2022), 404–429

  11. [11]

    Laura Carstensen, Derek Isaacowitz, and Susan Charles. 1999. Taking time seriously: A theory of socioemotional selectivity.The American psychologist54 (04 1999), 165–81. doi:10.1037//0003-066X.54.3.165

  12. [12]

    Dennett.The Intentional Stance

    Justine Cassell, Joseph Sullivan, Scott Prevost, and Elizabeth Churchill (Eds.). 2000.Embodied Conversational Agents. MIT Press, Cambridge, MA. https: //mitpress.mit.edu/9780262032787/embodied-conversational-agents/

  13. [13]

    Rijul Chaturvedi, Sanjeev Verma, Ronnie Das, and Yogesh K Dwivedi. 2023. Social companionship with artificial intelligence: Recent trends and future avenues. Technological Forecasting and Social Change193 (2023), 122634

  14. [14]

    Maximilian Croissant, Madeleine Frister, Guy Schofield, and Cade McCall. 2024. An appraisal-based chain-of-emotion architecture for affective language model game agents.PLOS ONE19 (05 2024), e0301033. doi:10.1371/journal.pone.0301033

  15. [15]

    Vedant Das Swain, Qiuyue "Joy" Zhong, Jash Rajesh Parekh, Yechan Jeon, Roy Zimmermann, Mary P Czerwinski, Jina Suh, Varun Mishra, Koustuv Saha, and Javier Hernandez. 2025. AI on My Shoulder: Supporting Emotional Labor in Front-Office Roles with an LLM-based Empathetic Coworker. InProceedings of the 2025 CHI Conference on Human Factors in Computing Systems...

  16. [16]

    Deci and Richard M

    Edward L. Deci and Richard M. Ryan. 1985.Education. Springer US, Boston, MA, 245–271. doi:10.1007/978-1-4899-2271-7_9

  17. [17]

    Raymond J. Dolan. 2002. Emotion, cognition, and behavior.Science298, 5596 (2002), 1191–1194. doi:10.1126/science.1076358

  18. [18]

    Zane Durante, Qiuyuan Huang, Naoki Wake, Ran Gong, Jae Sung Park, Bidipta Sarkar, Rohan Taori, Yusuke Noda, Demetri Terzopoulos, Yejin Choi, Katsushi Ikeuchi, Hoi Vo, Li Fei-Fei, and Jianfeng Gao. 2024. Agent AI: Surveying the Horizons of Multimodal Interaction. arXiv:2401.03568 [cs.AI] https://arxiv.org/ abs/2401.03568

  19. [19]

    Katherine Easton, Stephen Potter, Remi Bec, Matthew Bennion, Heidi Chris- tensen, Cheryl Grindell, Bahman Mirheidari, Scott Weich, Luc de Witte, Daniel Wolstenholme, et al. 2019. A virtual agent to support individuals living with physical and mental comorbidities: co-design and acceptability testing.Journal of medical Internet research21, 5 (2019), e12996

  20. [20]

    Daisy Fancourt, Henry Aughterson, Saoirse Finn, Emma Walker, and Andrew Steptoe. 2021. How leisure activities affect health: a narrative review and multi- level theoretical framework of mechanisms of action.The Lancet Psychiatry8, 4 (2021), 329–339. doi:10.1016/S2215-0366(20)30384-9

  21. [21]

    Cathy Mengying Fang, Phoebe Chua, Samantha W. T. Chan, Joanne Leong, Andria Bao, and Pattie Maes. 2025. Leveraging AI-Generated Emotional Self- Voice to Nudge People towards their Ideal Selves. InProceedings of the 2025 CHI Conference on Human Factors in Computing Systems (CHI ’25). Association for Computing Machinery, New York, NY, USA, Article 58, 20 pa...

  22. [22]

    How AI and Human Behaviors Shape Psychosocial Effects of Extended Chatbot Use: A Longitudinal Randomized Controlled Study

    Cathy Mengying Fang, Auren R. Liu, Valdemar Danry, Eunhae Lee, Samantha W. T. Chan, Pat Pataranutaporn, Pattie Maes, Jason Phang, Michael Lampe, Lama Ahmad, and Sandhini Agarwal. 2025. How AI and Human Behaviors Shape Psychosocial Effects of Extended Chatbot Use: A Longitudinal Randomized Controlled Study. arXiv:2503.17473 [cs.HC] https://arxiv.org/abs/2503.17473

  23. [23]

    Jodi Forlizzi and Katja Battarbee. 2004. Understanding experience in interactive systems. InProceedings of the 5th Conference on Designing Interactive Systems: Processes, Practices, Methods, and Techniques(Cambridge, MA, USA)(DIS ’04). Association for Computing Machinery, New York, NY, USA, 261–268. doi:10. 1145/1013115.1013152

  24. [24]

    N.H. Frijda. 1986.The Emotions. Cambridge University Press, The Triangle Building, Shaftesbury Road, Cambridge, CB2 8EA. https://books.google.com.sg/ books?id=QkNuuVf-pBMC

  25. [25]

    2011.Time-use surveys and the measurement of national well- being

    Jonathan Gershuny. 2011.Time-use surveys and the measurement of national well- being. Technical Report. Centre for Time Use Research, University of Oxford, Swansea, UK, Office for National Statistics

  26. [26]

    Jonathan Gruber, Eszter Hargittai, and Minh Hao Nguyen. 2022. The value of face-to-face communication in the digital world: What people miss about in-person interactions when those are limited.Studies in Communication Sciences 22 (10 2022). doi:10.24434/j.scoms.2022.03.3340

  27. [27]

    Guingrich and Michael S

    Rose E. Guingrich and Michael S. A. Graziano. 2025. A Longitudinal Randomized Control Study of Companion Chatbot Use: Anthropomorphism and Its Mediating Role on Social Impacts. arXiv:2509.19515 [cs.HC] https://arxiv.org/abs/2509. 19515

  28. [28]

    Hovland and Muzafer Sherif

    Carl I. Hovland and Muzafer Sherif. 1952. Judgment: Assimilation and Contrast Effects in Communication and Attitude Change.Journal of Abnormal and Social Psychology47, 3 (1952), 244–252. doi:10.1037/h0055775

  29. [29]

    Bo Hu, Yuanyi Mao, and Ki Joon Kim. 2023. How social anxiety leads to prob- lematic use of conversational AI: The roles of loneliness, rumination, and mind perception.Computers in Human Behavior145 (2023), 107760

  30. [30]

    Dongmei Hu, Yuting Lan, Haolan Yan, and Charles Weizheng Chen. 2025. What makes you attached to social companion AI? A two-stage exploratory mixed- method study.International Journal of Information Management83 (2025), 102890

  31. [31]

    Aaron Hurst, Adam Lerer, Adam P Goucher, Adam Perelman, Aditya Ramesh, Aidan Clark, AJ Ostrow, Akila Welihinda, Alan Hayes, Alec Radford, et al. 2024. Gpt-4o system card. arXiv:2410.21276 [cs.CL] https://arxiv.org/abs/2410.21276

  32. [32]

    Epstein, and Young-Ho Kim

    Eunkyung Jo, Yuin Jeong, Sohyun Park, Daniel A. Epstein, and Young-Ho Kim

  33. [33]

    In Proceedings of the 2024 CHI Conference on Human Factors in Computing Systems (Honolulu, HI, USA)(CHI ’24)

    Understanding the Impact of Long-Term Memory on Self-Disclosure with Large Language Model-Driven Chatbots for Public Health Intervention. In Proceedings of the 2024 CHI Conference on Human Factors in Computing Systems (Honolulu, HI, USA)(CHI ’24). Association for Computing Machinery, New York, NY, USA, Article 440, 21 pages. doi:10.1145/3613904.3642420

  34. [34]

    Yonghyun Jun and Hwanhee Lee. 2025. Exploring Persona Sentiment Sensitivity in Personalized Dialogue Generation. InProceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), Wanxiang Che, Joyce Nabende, Ekaterina Shutova, and Mohammad Taher Pilehvar (Eds.). Association for Computational Linguistics, Vi...

  35. [35]

    Jihyun Kim, Kelly Merrill Jr, Kun Xu, and Deanna D Sellnow. 2021. I like my relational machine teacher: An AI instructor’s communication styles and so- cial presence in online education.International Journal of Human–Computer Interaction37, 18 (2021), 1760–1770

  36. [36]

    Kyusik Kim, Hyungwoo Song, Jeongwoo Ryu, Changhoon Oh, and Bongwon Suh. 2025. BleacherBot: AI Agent as a Sports Co-Viewing Partner. InProceedings of the 2025 CHI Conference on Human Factors in Computing Systems (CHI ’25). Association for Computing Machinery, New York, NY, USA, Article 988, 31 pages. doi:10.1145/3706598.3714178

  37. [37]

    Alan B. Krueger. 2009. Introduction and Overview. InMeasuring the Subjective Well-Being of Nations: National Accounts of Time Use and Well-Being. University of Chicago Press, 1427 East 60th Street, Chicago, IL 60637 U.S.A. doi:10.7208/ chicago/9780226454573.003.0001

  38. [38]

    Seonghee Lee, Denae Ford, John Tang, Sasa Junuzovic, Asta Roseway, Ed Cutrell, and Kori Inkpen. 2025. IRL Dittos: Embodied Multimodal AI Agent Interactions in Open Spaces. arXiv:2504.21347 [cs.AI] https://arxiv.org/abs/2504.21347

  39. [39]

    Han Li and Renwen Zhang. 2024. Finding love in algorithms: deciphering the emotional contexts of close encounters with AI chatbots.Journal of Computer- Mediated Communication29, 5 (2024), zmae015

  40. [40]

    Tianjian Liu, Hongzheng Zhao, Yuheng Liu, Xingbo Wang, and Zhenhui Peng

  41. [41]

    InProceedings of the 37th Annual ACM Symposium on User Interface Software and Technology(Pittsburgh, PA, USA)(UIST ’24)

    ComPeer: A Generative Conversational Agent for Proactive Peer Support. InProceedings of the 37th Annual ACM Symposium on User Interface Software and Technology(Pittsburgh, PA, USA)(UIST ’24). Association for Computing Machinery, New York, NY, USA, Article 117, 22 pages. doi:10.1145/3654777. 3676430

  42. [42]

    Weber, Christopher K

    George Loewenstein, Elke U. Weber, Christopher K. Hsee, and Ned Welch. 2001. Risk as feelings.Journal of Economic Literature39, 2 (2001), 426–452. doi:10. 1257/jel.39.2.426

  43. [43]

    Graham Loomes and Robert Sugden. 1982. Regret theory: An alternative theory of rational choice under uncertainty.The Economic Journal92, 368 (1982), 805–

  44. [44]

    Bethanie Maples, Merve Cerit, Aditya Vishwanath, and Roy Pea. 2024. Loneliness and suicide mitigation for students using GPT3-enabled chatbots.npj Mental Health Research3 (01 2024). doi:10.1038/s44184-023-00047-6

  45. [45]

    Abraham Harold Maslow. 1943. A theory of human motivation.Psychological review50, 4 (1943), 370

  46. [46]

    Matud, Juan Bethencourt-Pére, M𝑎 Pino, D

    M. Matud, Juan Bethencourt-Pére, M𝑎 Pino, D. Estefanía Hernández-Lorenzo, Demelsa Fortes, and Ignacio Ibáñez. 2024. Time Use, Health, and Well-Being across the Life Cycle: A Gender Analysis.Social Sciences13 (06 2024), 307. doi:10.3390/socsci13060307

  47. [47]

    Mason McClay, Matthew E Sachs, and David Clewett. 2023. Dynamic emotional states shape the episodic structure of memory.Nature Communications14, 1 (2023), 6533

  48. [48]

    Mellers, Alan Schwartz, and Kahneman Ho

    Barbara A. Mellers, Alan Schwartz, and Kahneman Ho. 1997. Decision affect theory: Emotional reactions to the outcomes of risky options.Psychological Review104, 2 (1997), 267–285. doi:10.1037/0033-295X.104.2.267

  49. [49]

    Agata Mirowska and Jbid Arsenyan. 2023. Sweet escape: The role of empathy in social media engagement with human versus virtual influencers.International Journal of Human-Computer Studies174 (2023), 103008

  50. [50]

    Manisha Natarajan and Matthew Gombolay. 2020. Effects of Anthropomorphism and Accountability on Trust in Human Robot Interaction. InProceedings of the 2020 ACM/IEEE International Conference on Human-Robot Interaction(Cambridge, United Kingdom)(HRI ’20). Association for Computing Machinery, New York, NY, USA, 33–42. doi:10.1145/3319502.3374839

  51. [51]

    Humza Naveed, Asad Ullah Khan, Shi Qiu, Muhammad Saqib, Saeed Anwar, Muhammad Usman, Naveed Akhtar, Nick Barnes, and Ajmal Mian. 2025. A Comprehensive Overview of Large Language Models.ACM Trans. Intell. Syst. Technol.16, 5, Article 106 (Aug. 2025), 72 pages. doi:10.1145/3744746

  52. [52]

    Ochsner and Elizabeth A

    Kevin N. Ochsner and Elizabeth A. Phelps. 2007. Emerging perspectives on emotion–cognition interactions.Trends in Cognitive Sciences11, 8 (2007), 317–

  53. [53]

    doi:10.1016/j.tics.2007.06.008

  54. [54]

    O’Leary, and Timothy W

    Stefan Olafsson, Teresa K. O’Leary, and Timothy W. Bickmore. 2020. Motivating Health Behavior Change with Humorous Virtual Agents. InProceedings of the 20th ACM International Conference on Intelligent Virtual Agents(Virtual Event, Scotland, UK)(IV A ’20). Association for Computing Machinery, New York, NY, USA, Article 42, 8 pages. doi:10.1145/3383652.3423915

  55. [55]

    Ronan E O’Carroll and Martin S Hagger. 2013. Cognitive models of emotion and decision making.Neuroscience & Biobehavioral Reviews37, 8 (2013), 1866–1874

  56. [56]

    Jae Eun Park, Alei Fan, and Laurie Wu. 2025. Chatbots in complaint handling: The moderating role of humor.International Journal of Contemporary Hospitality Management37, 3 (2025), 805–824

  57. [57]

    , year = 2023, booktitle =

    Joon Sung Park, Joseph O’Brien, Carrie Jun Cai, Meredith Ringel Morris, Percy Liang, and Michael S. Bernstein. 2023. Generative Agents: Interactive Simulacra of Human Behavior. InProceedings of the 36th Annual ACM Symposium on User Interface Software and Technology(San Francisco, CA, USA)(UIST ’23). Association for Computing Machinery, New York, NY, USA, ...

  58. [58]

    My Boyfriend is AI

    Pat Pataranutaporn, Sheer Karny, Chayapatr Archiwaranguprok, Constanze Albrecht, Auren R. Liu, and Pattie Maes. 2025. "My Boyfriend is AI": A Com- putational Analysis of Human-AI Companionship in Reddit’s AI Community. https://api.semanticscholar.org/CorpusID:281315816 Qin et al

  59. [59]

    Pat Pataranutaporn, Kavin Winson, Peggy Yin, Auttasak Lapapirojn, Pichayoot Ouppaphan, Monchai Lertsutthiwong, Pattie Maes, and Hal Hershfield. 2024. Fu- ture You: A Conversation with an AI-Generated Future Self Reduces Anxiety, Neg- ative Emotions, and Increases Future Self-Continuity. arXiv:2405.12514 [cs.HC] https://arxiv.org/abs/2405.12514

  60. [60]

    Iryna Pentina, Tyler Hancock, and Tianling Xie. 2023. Exploring relationship development with social chatbots: A mixed-method study of replika.Computers in Human Behavior140 (2023), 107600

  61. [61]

    Luiz Pessoa. 2008. On the relationship between emotion and cognition.Nature Reviews Neuroscience9, 2 (2008), 148–158. doi:10.1038/nrn2317

  62. [62]

    Luiz Pessoa. 2009. How do emotion and motivation direct executive control? Trends in Cognitive Sciences13, 4 (2009), 160–166. doi:10.1016/j.tics.2009.01.006

  63. [63]

    Tiejun Qi, Hongshen Liu, and Zhihui Huang. 2025. An assistant or A friend? The role of parasocial relationship of human-computer interaction.Computers in Human Behavior167 (2025), 108625

  64. [64]

    Amon Rapp, Lorenzo Curti, and Arianna Boldi. 2021. The human side of human- chatbot interaction: A systematic literature review of ten years of research on text-based chatbots.International Journal of Human-Computer Studies151 (2021), 102630

  65. [65]

    Richard M Ryan and Edward L Deci. 2000. Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being.American psychologist55, 1 (2000), 68

  66. [66]

    Marita Skjuve, Asbjørn Følstad, Knut Inge Fostervold, and Petter Bae Brandtzaeg

  67. [67]

    My chatbot companion-a study of human-chatbot relationships.Interna- tional Journal of Human-Computer Studies149 (2021), 102601

  68. [68]

    Quan Tu, Chuanqi Chen, Jinpeng Li, Yanran Li, Shuo Shang, Dongyan Zhao, Ran Wang, and Rui Yan. 2023. CharacterChat: Learning towards Conversational AI with Personalized Social Support. arXiv:2308.10278 [cs.CL] https://arxiv.org/ abs/2308.10278

  69. [69]

    Morris, Gordon B

    Viswanath Venkatesh, Michael G. Morris, Gordon B. Davis, and Fred D. Davis

  70. [70]

    Management Information Systems Quarterly27, 3 (09 2003), 425–478

    User Acceptance of Information Technology: Toward A Unified View1. Management Information Systems Quarterly27, 3 (09 2003), 425–478. doi:10.2307/ 30036540

  71. [71]

    Lei Wang, Jingsen Zhang, Hao Yang, Zhi-Yuan Chen, Jiakai Tang, Zeyu Zhang, Xu Chen, Yankai Lin, Hao Sun, Ruihua Song, et al. 2025. User behavior simulation with large language model-based agents.ACM Transactions on Information Systems43, 2 (2025), 1–37

  72. [72]

    Allan Wigfield and Jacquelynne S Eccles. 2000. Expectancy–value theory of achievement motivation.Contemporary Educational Psychology25, 1 (2000), 68–81

  73. [73]

    Tianling Xie, Iryna Pentina, and Tyler Hancock. 2023. Friend, mentor, lover: does chatbot engagement lead to psychological dependence?Journal of service Management34, 4 (2023), 806–828

  74. [74]

    Yidan Yin, Nan Jia, and Cheryl J. Wakslak. 2024. AI can help people feel heard, but an AI label diminishes this impact.Proceedings of the National Academy of Sciences121, 14 (2024), e2319112121

  75. [75]

    Don’t look at me!

    Jingyi Yuan, Xixian Peng, Yichen Liu, and Qiuzhen Wang. 2025. “Don’t look at me!”: The role of avatars’ presentation style and gaze direction in social chatbot design.Computers in Human Behavior164 (2025), 108501

  76. [76]

    Marcel Zeelenberg. 1999. Anticipated regret, expected feedback and behavioral decision making.Journal of Consumer Psychology8, 2 (1999), 163–188. doi:10. 1207/s15327663jcp0802_01

  77. [77]

    Junbo Zhang, Jiandong Lu, Xiaolei Wang, Luning Liu, and Yuqiang Feng. 2024. Emotional expressions of care and concern by customer service chatbots: Im- proved customer attitudes despite perceived inauthenticity.Decision Support Systems186 (2024), 114314

  78. [78]

    Self-driven & personality-like

    Ruixue Zhaoyang, Martin Sliwinski, Lynn Martire, and Joshua Smyth. 2018. Age Differences in Adults’ Daily Social Interactions: An Ecological Momentary Assessment Study.Psychology and Aging33 (04 2018), 607–618. doi:10.1037/ pag0000242 A1 Appendix A1.1 Algorithm Flow and Prompts A1.1.1CTEM Status Representation Details.The CTEM status representation formal...