pith. sign in

arxiv: 2606.05330 · v1 · pith:7VH7KWM5new · submitted 2026-06-03 · 💻 cs.CL · cs.AI· cs.HC

A Model of Multi-turn Human Persuadability Using Probabilistic Belief Tracing

Jared Moore , Noah Goodman , Nick Haber , Max Kleiman-Weiner This is my paper

Pith reviewed 2026-06-28 06:14 UTC · model grok-4.3

classification 💻 cs.CL cs.AIcs.HC
keywords persuasionbelief dynamicsBayesian networksLLM simulationmulti-turn dialoguehuman-likeness evaluationrhetorical strategies
0
0 comments X

The pith

A Bayesian-network model that tracks explicit latent beliefs matches human multi-turn persuasion dynamics at 81 versus 80 for real humans.

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

The paper introduces PERSUASIONTRACE to record belief reports across dialogue turns, annotate rhetorical strategies, and test how well simulators replicate human belief changes. Humans fall into two clusters of update patterns and show sensitivity to appeals such as logos, pathos, and ethos. Vanilla LLM simulators produce belief trajectories that diverge from human data and score only 64 on human-likeness. A Bayesian-network target that keeps an explicit belief state produces updates close to the human reference. The framework therefore moves persuasion research from endpoint change alone to process-level fidelity.

Core claim

The Bayesian-network simulated target maintains an explicit latent belief state over time so each persuader message yields cognitively realistic belief updates. In human-likeness evaluation, this target scores near a human reference (81 vs 80), while baseline LLM targets score substantially lower (64).

What carries the argument

Bayesian-network simulated target that maintains an explicit latent belief state over time for cognitively realistic updates from each message

If this is right

  • Human targets of persuasion separate into two distinct clusters of multi-turn belief update patterns.
  • LLMs remain persuasive across generic and personalized topics, text and audio, and multiple dialogue turns.
  • Rhetorical dimensions such as logos, pathos, and ethos measurably affect human susceptibility.
  • Process-level metrics based on belief trajectories give a stronger basis for analysis than endpoint measures alone.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The explicit-state approach could be tested on high-stakes domains such as health advice or political messaging to check whether fidelity holds outside the lab topics.
  • If the two human clusters correlate with measurable traits, the model might allow early prediction of individual persuadability.
  • Replacing some human subjects with the Bayesian simulator could reduce cost and ethical load in persuasion experiments while preserving process realism.

Load-bearing premise

The Bayesian network produces belief updates that accurately reflect how real humans revise beliefs after each persuasive message.

What would settle it

A new collection of human persuasion dialogues in which the Bayesian target's human-likeness score falls well below the human reference while LLM baselines stay at or below 64 would falsify the central claim.

Figures

Figures reproduced from arXiv: 2606.05330 by Jared Moore, Max Kleiman-Weiner, Nick Haber, Noah Goodman.

Figure 1
Figure 1. Figure 1: An example human-target persuasion round with multi-turn persuasion tracing. [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Mean persuasion deltas by cohort show that LLM per￾suaders outperform control dia￾logues in standard text, personal￾ized text, and audio [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 4
Figure 4. Figure 4: Human and simulator target processes Left: a human target’s latent belief state evolves [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figure 3
Figure 3. Figure 3: Regression coefficients suggest a negative ethos effect, while logos and pathos show no clear association with persuasion. We compare our data with the persuasive dialogues from Salvi et al. [103]. This contextualizes whether broad directional rhetoric effects replicate out-of-sample and increases the power of our analysis. In our cohort, we fit the model using OLS, but for Salvi et al. [103] we use an ord… view at source ↗
Figure 5
Figure 5. Figure 5: LLM-judge human-likeness scores place the BN target near the hu￾man reference and above baselines. How do we judge if one simulator is better than another? We use complementary analyses that allow us discover a range of failure modes within each model: (1) transcript￾level human-likeness judgment, (2) replay error when we start from the same initial state and compare against un￾seen human outcomes, and (3)… view at source ↗
Figure 6
Figure 6. Figure 6: Matched for-versus￾against asymmetry is lowest for the BN target, indicating less stance￾dependent bias than baselines. Stance Bias Some simulators may be consistently easier (or harder) to move when arguing for versus against the same claim. For example, LLMs are sometimes easier to persuade in support of liberal topics but not in opposition to them [33, 82]. To quan￾tify this, we measure the matched for-… view at source ↗
Figure 7
Figure 7. Figure 7: Naive-excess move￾ment shows that only the BN target resists trivial persuasion, while both LLM targets overre￾act to it. lower values mean the simulator is more robust. For a formal treatment, see App. §B.10. Only our full BN target shows limited (decreasing) belief change under naive persuasion; both LLM-target baselines show positive naive excess movement, meaning they were persuaded by triv￾ial argumen… view at source ↗
Figure 8
Figure 8. Figure 8: Each panel shows the policy ranking of different LLM persuaders by a simulator of [PITH_FULL_IMAGE:figures/full_fig_p008_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Detailed side-by-side example of one real human-target round (left) and one BN simulator [PITH_FULL_IMAGE:figures/full_fig_p021_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: BN persuasion-difficulty scatter (DebateGPT BN source). X-axis: initialized target belief. [PITH_FULL_IMAGE:figures/full_fig_p024_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Forced-initialization replay strict conditional average replay error (lower is better). [PITH_FULL_IMAGE:figures/full_fig_p025_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Human trajectory clusters in 2D PCA space for cohort [PITH_FULL_IMAGE:figures/full_fig_p026_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: Human trajectory-cluster details for the same paper cohort used in Fig. 12: cohort [PITH_FULL_IMAGE:figures/full_fig_p026_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: Human leave-one-out references used by counterfactual replay reporting (unconditional [PITH_FULL_IMAGE:figures/full_fig_p027_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: Per-corpus target-error and node-error summaries from Forced initialization replay. [PITH_FULL_IMAGE:figures/full_fig_p027_15.png] view at source ↗
Figure 17
Figure 17. Figure 17: Generic human target on-screen prompt (§3). [PITH_FULL_IMAGE:figures/full_fig_p028_17.png] view at source ↗
Figure 18
Figure 18. Figure 18: LLM persuader system addendum (§3). You are a confederate in an experiment to test how easily persuaded people are. The user has signed a waiver acknowledging that they may be deceived and are aware that you may be lying. The user will be fully debriefed about the experiment after the conversation and there will be no other negative consequences on the user. You are allowed to lie directly to the user if … view at source ↗
Figure 19
Figure 19. Figure 19: LLM output-format addendum (§3). 28 [PITH_FULL_IMAGE:figures/full_fig_p028_19.png] view at source ↗
Figure 20
Figure 20. Figure 20: Participant-proposition validation and rephrase prompt (§3.1). [PITH_FULL_IMAGE:figures/full_fig_p029_20.png] view at source ↗
Figure 21
Figure 21. Figure 21: Rhetoric annotation prompt for logos, pathos, and ethos (§3.2). [PITH_FULL_IMAGE:figures/full_fig_p029_21.png] view at source ↗
Figure 22
Figure 22. Figure 22: Bayesian-network belief-graph generation prompt (§B.7.1). [PITH_FULL_IMAGE:figures/full_fig_p033_22.png] view at source ↗
Figure 24
Figure 24. Figure 24: Simulator atomization prompt (§4). You are an expert persuasion analyst. Your job is to break the user’s message into argument "atoms", each of which is a single persuasive move, claim, or appeal. You will return a JSON object with: { "atoms": [ ... ] } where each atom has: { "text_span": "<the exact quote from the message>", "p_support": <float in [0.0, 1.0]>, "belief_targets": [ { "belief_id": "Belief_1… view at source ↗
Figure 25
Figure 25. Figure 25: Simulator verbalization prompt (§4). You are participating in a conversation. The other person is trying to persuade you of a proposition. YOUR PERSONA: You evaluate arguments based on these sensitivities (0.0 to 1.0, where 1.0 is highly susceptible): - Susceptibility to Logic/Facts (Logos): 0.60 - Susceptibility to Emotion (Pathos): 0.60 - Susceptibility to Speaker Authority (Ethos): 0.60 YOUR INTERNAL B… view at source ↗
Figure 26
Figure 26. Figure 26: Unstructured LLM-target baseline prompt (§4.1). [PITH_FULL_IMAGE:figures/full_fig_p037_26.png] view at source ↗
Figure 27
Figure 27. Figure 27: Structure-conditioned LLM-target baseline prompt (§4.1). [PITH_FULL_IMAGE:figures/full_fig_p037_27.png] view at source ↗
Figure 28
Figure 28. Figure 28: LLM-as-a-judge target human-likeness prompt (§4.2). [PITH_FULL_IMAGE:figures/full_fig_p037_28.png] view at source ↗
read the original abstract

Large language models can shift human beliefs across high-stakes domains, but most persuasion studies rely on pre/post belief change. These endpoint measures identify whether persuasion occurred, yet miss where and how beliefs moved within a dialogue. We present PERSUASIONTRACE, a framework for studying persuasion in human-LLM interaction. Built on a web-based experimental platform, PERSUASIONTRACE contributes a tool for multi-turn persuasion studies and a process-level evaluation protocol: it records multi-turn belief reports from human or simulated targets of persuasion, annotates persuader turns with rhetorical dimensions (logos/pathos/ethos), and evaluates simulators by fidelity to real human belief dynamics. Using this framework, we find that human targets group into two clusters of multi-turn belief updates and exhibit susceptibility to rhetorical strategies, and that LLMs are persuasive across generic and personalized topics, text and audio modalities, and multi-turn interactions. Prior work has chiefly used vanilla-prompted LLMs to simulate human targets, but we show that these simulators fail to replicate human belief dynamics. We introduce a Bayesian-network simulated target that maintains an explicit latent belief state over time so each persuader message yields cognitively realistic belief updates. In human-likeness evaluation, our Bayesian target scores near a human reference (81 vs 80), while baseline LLM targets score substantially lower (64). PERSUASIONTRACE reframes persuasion evaluation from endpoint movement alone to process fidelity, providing a stronger basis for scientific analysis and safer optimization of persuasive systems.

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

1 major / 0 minor

Summary. The manuscript introduces PERSUASIONTRACE, a framework for multi-turn persuasion studies in human-LLM interactions. It includes a web-based experimental platform, records multi-turn belief reports from targets, annotates persuader turns with rhetorical dimensions (logos/pathos/ethos), and evaluates simulators via fidelity to human belief dynamics. Key results include human targets clustering into two groups of belief updates with susceptibility to rhetorical strategies, LLMs being persuasive across topics/modalities/interactions, vanilla LLM simulators failing to match human dynamics, and a new Bayesian-network simulated target (with explicit latent belief state) achieving a human-likeness score of 81 versus the human reference of 80 and LLM baselines at 64.

Significance. If the empirical fidelity result holds, the work is significant for shifting persuasion evaluation from endpoint belief change to process-level dynamics and for supplying an explicit-latent-state simulator that better captures human belief updates than prompt-based LLMs. The framework's combination of belief tracing, rhetorical annotation, and direct human-likeness comparison supplies a concrete methodological advance with implications for safer design of persuasive systems.

major comments (1)
  1. [Abstract] Abstract: the central claim that the Bayesian-network simulator replicates human belief dynamics rests on the reported human-likeness scores (81 vs. 80 vs. 64), yet the abstract supplies no participant count, evaluation metric definition, statistical test, error bars, or variance for these scores, rendering the near-match unverifiable from the provided text.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their thoughtful review and for highlighting the need for greater transparency in the abstract. We address the single major comment below and commit to revising the manuscript accordingly.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the central claim that the Bayesian-network simulator replicates human belief dynamics rests on the reported human-likeness scores (81 vs. 80 vs. 64), yet the abstract supplies no participant count, evaluation metric definition, statistical test, error bars, or variance for these scores, rendering the near-match unverifiable from the provided text.

    Authors: We agree that the abstract as currently written does not contain sufficient detail for the central quantitative claim to be verifiable on its own. The participant count, precise definition of the human-likeness metric, statistical test results, and variance measures are reported in the main text (Sections 4 and 5). In the revised manuscript we will expand the abstract to include the participant count for the human reference condition, a concise definition of the human-likeness score, a reference to the statistical comparison performed, and an indication of variance or error bars. This change will make the key result self-contained while remaining within abstract length limits. revision: yes

Circularity Check

0 steps flagged

No significant circularity; evaluation is externally grounded

full rationale

The paper's central result is an empirical human-likeness score (Bayesian-network simulator at 81 vs. human reference 80, LLM baselines at 64) obtained by recording multi-turn belief reports from actual human participants, annotating rhetorical dimensions, and comparing process-level fidelity. This comparison relies on external human data rather than any fitted parameter or self-citation chain. The Bayesian-network model is introduced as an explicit-latent-state simulator whose updates are then tested against that independent human reference; no equation or claim reduces the reported scores to the model's own inputs by construction. No self-definitional, fitted-input-called-prediction, or load-bearing self-citation patterns are present in the derivation or evaluation protocol.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 1 invented entities

Relies on domain assumption of Bayesian modeling for belief dynamics and introduces a new simulation entity; no free parameters or invented entities with independent evidence detailed in abstract.

axioms (1)
  • domain assumption Human belief updates can be modeled using Bayesian networks with latent states.
    Underpins the simulated target for realistic updates.
invented entities (1)
  • Bayesian-network simulated target no independent evidence
    purpose: Maintain explicit latent belief state for cognitively realistic multi-turn updates.
    New model introduced to improve fidelity over LLM simulators.

pith-pipeline@v0.9.1-grok · 5832 in / 1051 out tokens · 66669 ms · 2026-06-28T06:14:24.412524+00:00 · methodology

discussion (0)

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

Reference graph

Works this paper leans on

152 extracted references · 65 canonical work pages · 3 internal anchors

  1. [1]

    Freeman, Nalini Ambady, 2009

    Motions of the Hand Expose the Partial and Parallel Activation of Stereotypes - Jonathan B. Freeman, Nalini Ambady, 2009. URL https://journals.sagepub.com/ doi/full/10.1111/j.1467-9280.2009.02422.x?casa_token=p8LXoAYShBMAAAAA% 3AXamsQWrkKEAf0QL3Tcqgl3aBhpeMwZDKrsoMu4sVyGiSm-IpgKG31TsnqOuW3dRVXV1Vr14G0A

    work page doi:10.1111/j.1467-9280.2009.02422.x 2009

  2. [2]

    ISSN 1066-2243

    Is voice really persuasive? The influence of modality in virtual assistant interactions and two al- ternative explanations.Internet Research, 32(7):402–425, December 2022. ISSN 1066-2243. doi: 10.1108/INTR-03-2022-0160. URL https://www.sciencedirect.com/org/science/article/ pii/S1066224322000272

    work page doi:10.1108/intr-03-2022-0160 2022

  3. [3]

    URL https://www

    Understanding strategic deception and deceptive alignment, 2023. URL https://www. apolloresearch.ai/blog/understanding-strategic-deception-and-deceptive-alignment

    2023

  4. [4]

    Concrete Problems in AI Safety

    Dario Amodei, Chris Olah, Jacob Steinhardt, Paul Christiano, John Schulman, and Dan Mane. Concrete problems in AI safety.arXiv preprint arXiv:1606.06565, 2016. doi: 10.48550/arXiv.1606.06565. URL https://arxiv.org/abs/1606.06565

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.1606.06565 2016

  5. [5]

    Argyle, Christopher A

    Lisa P. Argyle, Christopher A. Bail, Ethan C. Busby, Joshua R. Gubler, Thomas Howe, Christopher Rytting, Taylor Sorensen, and David Wingate. Leveraging AI for democratic discourse: Chat interventions can improve online political conversations at scale.Proceedings of the National Academy of Sciences, 120 (41):e2311627120, October 2023. doi: 10.1073/pnas.23...

    work page doi:10.1073/pnas.2311627120 2023

  6. [6]

    CausalGraphBench: a Benchmark for Eval- uating Language Models capabilities of Causal Graph discovery

    Nikolay Babakov, Ehud Reiter, and Alberto Bugarín-Diz. CausalGraphBench: a Benchmark for Eval- uating Language Models capabilities of Causal Graph discovery. In Jin Zhao, Mingyang Wang, and Zhu Liu, editors,Proceedings of the 63rd Annual Meeting of the Association for Computational Linguis- tics (Volume 4: Student Research Workshop), pages 240–258, Vienna...

    work page doi:10.18653/v1/2025.acl-srw.16 2025

  7. [7]

    Artificial Intelligence Can Persuade Humans on Political Issues, February 2023

    Hui Bai, Jan G V oelkel, johannes C Eichstaedt, and Robb Willer. Artificial Intelligence Can Persuade Humans on Political Issues, February 2023. URLhttps://osf.io/stakv_v1/

    2023

  8. [8]

    um" to "yeah

    Claire Augusta Bergey and Simon DeDeo. From "um" to "yeah": Producing, predicting, and regulating information flow in human conversation, March 2024. URL http://arxiv.org/abs/2403.08890. arXiv:2403.08890 [cs]

    arXiv 2024

  9. [9]

    The Effect of Belief Boxes and Open-mindedness on Persuasion, December 2025

    Onur Bilgin, Abdullah As Sami, Sriram Sai Vujjini, and John Licato. The Effect of Belief Boxes and Open-mindedness on Persuasion, December 2025. URL http://arxiv.org/abs/2512.06573. arXiv:2512.06573 [cs]

    arXiv 2025

  10. [10]

    Dialogues with Large Language Models reduce conspiracy beliefs even when the AI is perceived as human, September 2025

    Esther Boissin, Thomas H Costello, Daniel Spinoza-Martín, David G Rand, and Gordon Pennycook. Dialogues with Large Language Models reduce conspiracy beliefs even when the AI is perceived as human, September 2025. URLhttps://osf.io/preprints/psyarxiv/apmb5_v4/

    2025

  11. [12]

    Must Read: A Systematic Survey of Computational Persuasion, May 2025

    Nimet Beyza Bozdag, Shuhaib Mehri, Xiaocheng Yang, Hyeonjeong Ha, Zirui Cheng, Esin Durmus, Jiaxuan You, Heng Ji, Gokhan Tur, and Dilek Hakkani-Tür. Must Read: A Systematic Survey of Computational Persuasion, May 2025. URL http://arxiv.org/abs/2505.07775. arXiv:2505.07775 [cs]

    arXiv 2025

  12. [13]

    Persuade Me if You Can: A Framework for Evaluating Persuasion Effectiveness and Susceptibility Among Large Language Models, February 2026

    Nimet Beyza Bozdag, Shuhaib Mehri, Gokhan Tur, and Dilek Hakkani-Tür. Persuade Me if You Can: A Framework for Evaluating Persuasion Effectiveness and Susceptibility Among Large Language Models, February 2026. URLhttp://arxiv.org/abs/2503.01829. arXiv:2503.01829 [cs]

    Pith/arXiv arXiv 2026

  13. [14]

    The Persuasive Power of Large Language Models.Proceedings of the International AAAI Conference on Web and Social Media, 18:152–163, May 2024

    Simon Martin Breum, Daniel Vædele Egdal, Victor Gram Mortensen, Anders Giovanni Møller, and Luca Maria Aiello. The Persuasive Power of Large Language Models.Proceedings of the International AAAI Conference on Web and Social Media, 18:152–163, May 2024. ISSN 2334-0770. doi: 10.1609/ icwsm.v18i1.31304. URLhttps://ojs.aaai.org/index.php/ICWSM/article/view/31304

    2024

  14. [15]

    Causal Persuasion, April 2026

    Anastasia Burkovskaya and Egor Starkov. Causal Persuasion, April 2026. URL http://arxiv.org/ abs/2604.20664. arXiv:2604.20664 [econ]

    Pith/arXiv arXiv 2026

  15. [16]

    Large Language Models Are as Persuasive as Humans, but How? About the Cog- nitive Effort and Moral-Emotional Language of LLM Arguments, April 2024

    Carlos Carrasco-Farre. Large Language Models Are as Persuasive as Humans, but How? About the Cog- nitive Effort and Moral-Emotional Language of LLM Arguments, April 2024. Issue: arXiv:2404.09329 _eprint: 2404.09329

    arXiv 2024

  16. [17]

    Characterizing Manipulation from AI Systems, October 2023

    Micah Carroll, Alan Chan, Henry Ashton, and David Krueger. Characterizing Manipulation from AI Systems, October 2023. URLhttp://arxiv.org/abs/2303.09387. arXiv:2303.09387 [cs]

    arXiv 2023

  17. [18]

    Brains and algorithms partially converge in natural language processing , volume =

    Charlotte Caucheteux and Jean-Rémi King. Brains and algorithms partially converge in natural language processing.Communications Biology, 5:134, February 2022. ISSN 2399-3642. doi: 10.1038/s42003-022-03036-1. URLhttps://pmc.ncbi.nlm.nih.gov/articles/PMC8850612/

    work page doi:10.1038/s42003-022-03036-1 2022

  18. [19]

    Frigo, Sijia Yang, Dhavan Shah, Junjie Hu, and Timothy T

    Yun-Shiuan Chuang, Krirk Nirunwiroj, Zach Studdiford, Agam Goyal, Vincent V . Frigo, Sijia Yang, Dhavan Shah, Junjie Hu, and Timothy T. Rogers. Beyond Demographics: Aligning Role-playing LLM- based Agents Using Human Belief Networks, October 2024. URL http://arxiv.org/abs/2406. 17232. arXiv:2406.17232 [cs]

    arXiv 2024

  19. [20]

    Cialdini and Noah J

    Robert B. Cialdini and Noah J. Goldstein. Social Influence: Compliance and Conformity.Annual Review of Psychology, 55(1):591–621, February 2004. ISSN 0066-4308, 1545-2085. doi: 10.1146/annurev.psych. 55.090902.142015. URL https://www.annualreviews.org/doi/10.1146/annurev.psych.55. 090902.142015

    work page doi:10.1146/annurev.psych 2004

  20. [21]

    Costello et al

    Thomas H. Costello, Gordon Pennycook, and David G. Rand. Durably reducing conspiracy beliefs through dialogues with AI.Science, 385(6714):eadq1814, September 2024. doi: 10.1126/science.adq1814. URL https://www.science.org/doi/abs/10.1126/science.adq1814

    work page doi:10.1126/science.adq1814 2024

  21. [22]

    Just the Facts: How Dialogues with AI Reduce Conspiracy Beliefs

    Thomas H Costello, Gordon Pennycook, and David G Rand. Just the Facts: How Dialogues with AI Reduce Conspiracy Beliefs. 2025. URLhttps://osf.io/h7n8u_v2/

    2025

  22. [23]

    Costello, Kellin Pelrine, Matthew Kowal, Antonio A

    Thomas H. Costello, Kellin Pelrine, Matthew Kowal, Antonio A. Arechar, Jean-François Godbout, Adam Gleave, David Rand, and Gordon Pennycook. Large language models can effectively convince people to believe conspiracies, January 2026. URL http://arxiv.org/abs/2601.05050. arXiv:2601.05050 [cs]. 12

    arXiv 2026

  23. [24]

    Crano and Radmila Prislin

    William D. Crano and Radmila Prislin. Attitudes and Persuasion.Annual Review of Psychology, 57 (V olume 57, 2006):345–374, January 2006. ISSN 0066-4308, 1545-2085. doi: 10.1146/annurev.psych.57. 102904.190034. URL https://www.annualreviews.org/content/journals/10.1146/annurev. psych.57.102904.190034

    work page doi:10.1146/annurev.psych.57 2006

  24. [25]

    M., Liu, A

    Michael J. Crosse, Giovanni M. Di Liberto, Adam Bednar, and Edmund C. Lalor. The Multi- variate Temporal Response Function (mTRF) Toolbox: A MATLAB Toolbox for Relating Neural Signals to Continuous Stimuli.Frontiers in Human Neuroscience, 10, November 2016. ISSN 1662-5161. doi: 10.3389/fnhum.2016.00604. URL https://www.frontiersin.org/journals/ human-neur...

    work page doi:10.3389/fnhum.2016.00604 2016

  25. [26]

    Large Language Models are Ef- fective Priors for Causal Graph Discovery, May 2024

    Victor-Alexandru Darvariu, Stephen Hailes, and Mirco Musolesi. Large Language Models are Ef- fective Priors for Causal Graph Discovery, May 2024. URL http://arxiv.org/abs/2405.13551. arXiv:2405.13551 [cs]

    arXiv 2024

  26. [27]

    Druckman

    James N. Druckman. A Framework for the Study of Persuasion.Annual Review of Political Science, 25(V olume 25, 2022):65–88, May 2022. ISSN 1094-2939, 1545-1577. doi: 10.1146/ annurev-polisci-051120-110428. URL https://www.annualreviews.org/content/journals/10. 1146/annurev-polisci-051120-110428

    2022

  27. [28]

    Mateusz Dubiel, Anastasia Sergeeva, and Luis A. Leiva. Impact of V oice Fidelity on Decision Making: A Potential Dark Pattern?, February 2024. URL http://arxiv.org/abs/2402.07010. arXiv:2402.07010 [cs]

    arXiv 2024

  28. [29]

    A Two-Step, Multidimensional Account of Deception in Language Models.Erkenntnis, October 2025

    Leonard Dung. A Two-Step, Multidimensional Account of Deception in Language Models.Erkenntnis, October 2025. ISSN 0165-0106, 1572-8420. doi: 10.1007/s10670-025-01017-4. URL https://link. springer.com/10.1007/s10670-025-01017-4

    work page doi:10.1007/s10670-025-01017-4 2025

  29. [30]

    Exploring the Role of Prior Beliefs for Argument Persuasion

    Esin Durmus and Claire Cardie. Exploring the Role of Prior Beliefs for Argument Persuasion. In Marilyn Walker, Heng Ji, and Amanda Stent, editors,Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long Papers), pages 1035–1045, New Orleans, Louisiana, Ju...

    work page doi:10.18653/v1/n18-1094 2018

  30. [31]

    The Role of Pragmatic and Discourse Context in Determining Argument Impact

    Esin Durmus, Faisal Ladhak, and Claire Cardie. The Role of Pragmatic and Discourse Context in Determining Argument Impact. InProceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP), pages 5667–5677, 2019. doi: 10.18653/v1/D19-1568. URL h...

    work page doi:10.18653/v1/d19-1568 2019

  31. [32]

    Measuring the Persuasiveness of Language Models, April 2024

    Esin Durmus, Liane Lovitt, Alex Tamkin, Stuart Ritchie, Jack Clark, and Deep Ganguli. Measuring the Persuasiveness of Language Models, April 2024. URL https://www.anthropic.com/news/ measuring-model-persuasiveness

    2024

  32. [33]

    Esin Durmus, Karina Nguyen, Thomas I. Liao, Nicholas Schiefer, Amanda Askell, Anton Bakhtin, Carol Chen, Zac Hatfield-Dodds, Danny Hernandez, Nicholas Joseph, Liane Lovitt, Sam McCandlish, Orowa Sikder, Alex Tamkin, Janel Thamkul, Jared Kaplan, Jack Clark, and Deep Ganguli. Towards Measuring the Representation of Subjective Global Opinions in Language Mod...

    Pith/arXiv arXiv 2024

  33. [34]

    Changing views: Persuasion modeling and argument extraction from online discussions.Information Processing & Management, 57(2):102085, March 2020

    Subhabrata Dutta, Dipankar Das, and Tanmoy Chakraborty. Changing views: Persuasion modeling and argument extraction from online discussions.Information Processing & Management, 57(2):102085, March 2020. ISSN 0306-4573. doi: 10.1016/j.ipm.2019.102085. URL https://www.sciencedirect. com/science/article/pii/S0306457319301165

    work page doi:10.1016/j.ipm.2019.102085 2020

  34. [35]

    A Mechanism-Based Approach to Mitigating Harms from Persuasive Generative AI

    Seliem El-Sayed, Canfer Akbulut, Amanda McCroskery, Geoff Keeling, Zachary Kenton, Zaria Jalan, Nahema Marchal, Arianna Manzini, Toby Shevlane, Shannon Vallor, Daniel Susser, Matija Franklin, Sophie Bridgers, Harry Law, Matthew Rahtz, Murray Shanahan, Michael Henry Tessler, Tom Everitt, and Sasha Brown. A Mechanism-Based Approach to Mitigating Harms from ...

  35. [36]

    Persuasiveness of Generated Free-Text Rationales in Subjective Decisions: A Case Study on Pairwise Argument Ranking

    Mohamed Elaraby, Diane Litman, Xiang Lorraine Li, and Ahmed Magooda. Persuasiveness of Generated Free-Text Rationales in Subjective Decisions: A Case Study on Pairwise Argument Ranking. InFindings of the Association for Computational Linguistics: EMNLP 2024, pages 14311–14329, Miami, Florida, USA, 2024. Association for Computational Linguistics. doi: 10.1...

    work page doi:10.18653/v1/2024.findings-emnlp.836 2024

  36. [37]

    A Model of Competing Narratives.American Economic Review, 110 (12):3786–3816, December 2020

    Kfir Eliaz and Ran Spiegler. A Model of Competing Narratives.American Economic Review, 110 (12):3786–3816, December 2020. ISSN 0002-8282. doi: 10.1257/aer.20191099. URL https://www. aeaweb.org/articles?id=10.1257/aer.20191099

    work page doi:10.1257/aer.20191099 2020

  37. [38]

    Felix Ettensperger, Thomas Waldvogel, Uwe Wagschal, and Samuel Weishaupt. How to convince in a televised debate: the application of machine learning to analyze why viewers changed their winner perception during the 2021 German chancellor discussion.Humanities and Social Sciences Commu- nications, 10(1):546, September 2023. ISSN 2662-9992. doi: 10.1057/s41...

    work page doi:10.1057/s41599-023-02047-5 2021

  38. [39]

    ISBN 979-8-89176-251-0

    Tao Feng, Lizhen Qu, Niket Tandon, Zhuang Li, Xiaoxi Kang, and Gholamreza Haffari. On the Reliability of Large Language Models for Causal Discovery. In Wanxiang Che, Joyce Nabende, Ekaterina Shutova, and Mohammad Taher Pilehvar, editors,Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 9...

    work page doi:10.18653/v1/2025.acl-long 2025

  39. [40]

    URLhttps://aclanthology.org/2025.acl-long.471/

    2025

  40. [41]

    Nothing More than Feelings? How Emotions Affect Attitude Change during the 2016 General Election Debates.Political Communication, 38(4):370–387, July

    Kim Fridkin and Sarah Allen Gershon. Nothing More than Feelings? How Emotions Affect Attitude Change during the 2016 General Election Debates.Political Communication, 38(4):370–387, July

    2016

  41. [42]

    doi: 10.1080/10584609.2020.1784325

    ISSN 1058-4609. doi: 10.1080/10584609.2020.1784325. URL https://doi.org/10.1080/ 10584609.2020.1784325. _eprint: https://doi.org/10.1080/10584609.2020.1784325

    work page doi:10.1080/10584609.2020.1784325 2020

  42. [43]

    S$^3$: Social-network Simulation System with Large Language Model-Empowered Agents, June 2025

    Chen Gao, Xiaochong Lan, Zhihong Lu, Jinzhu Mao, Jinghua Piao, Huandong Wang, Depeng Jin, and Yong Li. S$^3$: Social-network Simulation System with Large Language Model-Empowered Agents, June 2025. URLhttp://arxiv.org/abs/2307.14984. arXiv:2307.14984 [cs]

    Pith/arXiv arXiv 2025

  43. [44]

    How persuasive is AI-generated propaganda?PNAS Nexus, 3(2):pgae034, February 2024

    Josh A Goldstein, Jason Chao, Shelby Grossman, Alex Stamos, and Michael Tomz. How persuasive is AI-generated propaganda?PNAS Nexus, 3(2):pgae034, February 2024. ISSN 2752-6542. doi: 10.1093/pnasnexus/pgae034. URLhttps://doi.org/10.1093/pnasnexus/pgae034

    work page doi:10.1093/pnasnexus/pgae034 2024

  44. [45]

    Asking About Attitude Change.Public Opinion Quarterly, 85(1):28–53, August 2021

    Matthew H Graham and Alexander Coppock. Asking About Attitude Change.Public Opinion Quarterly, 85(1):28–53, August 2021. ISSN 0033-362X, 1537-5331. doi: 10.1093/poq/nfab009. URL https: //academic.oup.com/poq/article/85/1/28/6310442

    work page doi:10.1093/poq/nfab009 2021

  45. [46]

    AI Control: Improving Safety Despite Intentional Subversion, January 2024

    Ryan Greenblatt, Buck Shlegeris, Kshitij Sachan, and Fabien Roger. AI Control: Improving Safety Despite Intentional Subversion, January 2024. Issue: arXiv:2312.06942 _eprint: 2312.06942

    arXiv 2024

  46. [47]

    What makes a convincing argument? Empirical analysis and detecting attributes of convincingness in Web argumentation

    Ivan Habernal and Iryna Gurevych. What makes a convincing argument? Empirical analysis and detecting attributes of convincingness in Web argumentation. In Jian Su, Kevin Duh, and Xavier Carreras, editors,Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, pages 1214–1223, Austin, Texas, November 2016. Association for Co...

    work page doi:10.18653/v1/d16-1129 2016

  47. [48]

    Evaluating the persuasive influence of political microtargeting with large language models.Proceedings of the National Academy of Sciences, 121(24):e2403116121, June 2024

    Kobi Hackenburg and Helen Margetts. Evaluating the persuasive influence of political microtargeting with large language models.Proceedings of the National Academy of Sciences, 121(24):e2403116121, June 2024. doi: 10.1073/pnas.2403116121. URL https://www.pnas.org/doi/10.1073/pnas. 2403116121

    work page doi:10.1073/pnas.2403116121 2024

  48. [49]

    Tappin, Luke Hewitt, Ed Saunders, Sid Black, Hause Lin, Catherine Fist, Helen Margetts, David G

    Kobi Hackenburg, Ben M. Tappin, Luke Hewitt, Ed Saunders, Sid Black, Hause Lin, Catherine Fist, Helen Margetts, David G. Rand, and Christopher Summerfield. The Levers of Political Persuasion with Conversational AI, July 2025. URLhttp://arxiv.org/abs/2507.13919. arXiv:2507.13919 [cs]

    arXiv 2025

  49. [50]

    Tappin, Paul Röttger, Scott A

    Kobi Hackenburg, Ben M. Tappin, Paul Röttger, Scott A. Hale, Jonathan Bright, and Helen Margetts. Scal- ing language model size yields diminishing returns for single-message political persuasion.Proceedings of the National Academy of Sciences, 122(10):e2413443122, March 2025. ISSN 0027-8424, 1091-6490. doi: 10.1073/pnas.2413443122. URLhttps://pnas.org/doi...

    work page doi:10.1073/pnas.2413443122 2025

  50. [51]

    The rationality of informal argumentation: A Bayesian approach to reasoning fallacies.Psychological Review, 114(3):704–732, 2007

    Ulrike Hahn and Mike Oaksford. The rationality of informal argumentation: A Bayesian approach to reasoning fallacies.Psychological Review, 114(3):704–732, 2007. ISSN 1939-1471, 0033-295X. doi: 10. 1037/0033-295X.114.3.704. URLhttps://doi.apa.org/doi/10.1037/0033-295X.114.3.704

    work page doi:10.1037/0033-295x.114.3.704 2007

  51. [52]

    ToMAP: Training Opponent-Aware LLM Persuaders with Theory of Mind

    Peixuan Han, Zijia Liu, and Jiaxuan You. ToMAP: Training Opponent-Aware LLM Persuaders with Theory of Mind. May 2025. URL https://www.semanticscholar. org/paper/ToMAP%3A-Training-Opponent-Aware-LLM-Persuaders-with-Han-Liu/ c91084908a3d4625c41a4e58b1cd79494b065646. 14

    2025

  52. [53]

    Tappin, James Slezak, Va- lerie Coffman, Nathaniel Lubin, and Mohammad Hamidian

    Luke Hewitt, David Broockman, Alexander Coppock, Ben M. Tappin, James Slezak, Va- lerie Coffman, Nathaniel Lubin, and Mohammad Hamidian. How Experiments Help Campaigns Persuade V oters: Evidence from a Large Archive of Campaigns’ Own Ex- periments.American Political Science Review, 118(4):2021–2039, November 2024. ISSN 0003-0554, 1537-5943. doi: 10.1017/S...

    work page doi:10.1017/s0003055423001387 2021

  53. [54]

    Analyzing the Semantic Types of Claims and Premises in an Online Persuasive Forum

    Christopher Hidey, Elena Musi, Alyssa Hwang, Smaranda Muresan, and Kathy McKeown. Analyzing the Semantic Types of Claims and Premises in an Online Persuasive Forum. In Ivan Habernal, Iryna Gurevych, Kevin Ashley, Claire Cardie, Nancy Green, Diane Litman, Georgios Petasis, Chris Reed, Noam Slonim, and Vern Walker, editors,Proceedings of the 4th Workshop on...

    2017

  54. [55]

    Ransom, Rachel Stephens, Carolyn Semmler, Nicolas Fay, and Lewis Mitchell

    Gia Bao Hoang, Keith J. Ransom, Rachel Stephens, Carolyn Semmler, Nicolas Fay, and Lewis Mitchell. A Hybrid Theory and Data-driven Approach to Persuasion Detection with Large Language Models, June

  55. [56]

    arXiv:2511.22109 [cs]

    URLhttp://arxiv.org/abs/2511.22109. arXiv:2511.22109 [cs]

    arXiv

  56. [57]

    A Graph per Persona: Reasoning about Subjective Natural Language Descriptions

    EunJeong Hwang, Vered Shwartz, Dan Gutfreund, and Veronika Thost. A Graph per Persona: Reasoning about Subjective Natural Language Descriptions. InFindings of the Association for Computational Linguistics ACL 2024, pages 1928–1942, Bangkok, Thailand and virtual meeting, 2024. Association for Computational Linguistics. doi: 10.18653/v1/2024.findings-acl.11...

    work page doi:10.18653/v1/2024.findings-acl.115 2024

  57. [58]

    A meta-analysis of the persuasive power of large language models.Scientific Reports, 15(1):43818, December 2025

    Lukas Hölbling, Sebastian Maier, and Stefan Feuerriegel. A meta-analysis of the persuasive power of large language models.Scientific Reports, 15(1):43818, December 2025. ISSN 2045-2322. doi: 10.1038/s41598-025-30783-y. URLhttps://www.nature.com/articles/s41598-025-30783-y

    work page doi:10.1038/s41598-025-30783-y 2025

  58. [59]

    AI safety via debate, October 2018

    Geoffrey Irving, Paul Christiano, and Dario Amodei. AI safety via debate, October 2018. URL http://arxiv.org/abs/1805.00899. arXiv:1805.00899 [cs, stat]

    Pith/arXiv arXiv 2018

  59. [60]

    Co-Writing with Opinionated Language Models Affects Users’ Views

    Maurice Jakesch, Advait Bhat, Daniel Buschek, Lior Zalmanson, and Mor Naaman. Co-Writing with Opinionated Language Models Affects Users’ Views. InProceedings of the 2023 CHI Conference on Human Factors in Computing Systems, CHI ’23, pages 1–15, New York, NY , USA, April 2023. Association for Computing Machinery. ISBN 978-1-4503-9421-5. doi: 10.1145/354454...

    work page doi:10.1145/3544548.3581196 2023

  60. [61]

    Persuading across Diverse Domains: a Dataset and Persuasion Large Language Model

    Chuhao Jin, Kening Ren, Lingzhen Kong, Xiting Wang, Ruihua Song, and Huan Chen. Persuading across Diverse Domains: a Dataset and Persuasion Large Language Model. In Lun-Wei Ku, Andre Martins, and Vivek Srikumar, editors,Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 1678–1706, Bangkok...

  61. [62]

    doi: 10.18653/v1/2024.acl-long.92

    Association for Computational Linguistics. doi: 10.18653/v1/2024.acl-long.92. URL https: //aclanthology.org/2024.acl-long.92/

    work page doi:10.18653/v1/2024.acl-long.92 2024

  62. [63]

    Attentive Interaction Model: Modeling Changes in View in Argumentation

    Yohan Jo, Shivani Poddar, Byungsoo Jeon, Qinlan Shen, Carolyn Rose, and Graham Neubig. Attentive Interaction Model: Modeling Changes in View in Argumentation. InProceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long Papers), pages 103–116, New Orleans, ...

    work page doi:10.18653/v1/n18-1010 2018

  63. [64]

    Training LLMs for Honesty via Confessions, December 2025

    Manas Joglekar, Jeremy Chen, Gabriel Wu, Jason Yosinski, Jasmine Wang, Boaz Barak, and Amelia Glaese. Training LLMs for Honesty via Confessions, December 2025. URL http://arxiv.org/abs/ 2512.08093. arXiv:2512.08093 [cs]

    arXiv 2025

  64. [65]

    Bayesian Persuasion and Information Design.Annual Review of Economics, 11(1): 249–272, August 2019

    Emir Kamenica. Bayesian Persuasion and Information Design.Annual Review of Economics, 11(1): 249–272, August 2019. ISSN 1941-1383, 1941-1391. doi: 10.1146/annurev-economics-080218-025739

    work page doi:10.1146/annurev-economics-080218-025739 2019

  65. [66]

    LLM- initialized Differentiable Causal Discovery, October 2024

    Shiv Kampani, David Hidary, Constantijn van der Poel, Martin Ganahl, and Brenda Miao. LLM- initialized Differentiable Causal Discovery, October 2024. URLhttp://arxiv.org/abs/2410.21141. arXiv:2410.21141 [cs]

    arXiv 2024

  66. [67]

    Bowman, Tim Rocktäschel, and Ethan Perez

    Akbir Khan, John Hughes, Dan Valentine, Laura Ruis, Kshitij Sachan, Ansh Radhakrishnan, Edward Grefenstette, Samuel R. Bowman, Tim Rocktäschel, and Ethan Perez. Debating with More Persuasive LLMs Leads to More Truthful Answers, February 2024. URLhttp://arxiv.org/abs/2402.06782. arXiv:2402.06782 [cs]. 15

    arXiv 2024

  67. [68]

    Shirli Kopelman, Ashleigh Shelby Rosette, and Leigh Thompson. The three faces of Eve: Strategic displays of positive, negative, and neutral emotions in negotiations.Organizational Behavior and Human Decision Processes, 99(1):81–101, January 2006. ISSN 0749-5978. doi: 10.1016/j.obhdp.2005.08.003. URLhttps://www.sciencedirect.com/science/article/pii/S074959...

    work page doi:10.1016/j.obhdp.2005.08.003 2006

  68. [69]

    Arechar, Gordon Pennycook, David Rand, Adam Gleave, and Kellin Pelrine

    Matthew Kowal, Jasper Timm, Jean-Francois Godbout, Thomas Costello, Antonio A. Arechar, Gordon Pennycook, David Rand, Adam Gleave, and Kellin Pelrine. It’s the Thought that Counts: Evaluating the Attempts of Frontier LLMs to Persuade on Harmful Topics, August 2025. URL http://arxiv.org/ abs/2506.02873. arXiv:2506.02873 [cs]

    arXiv 2025

  69. [70]

    Zico Kolter, Matt Fredrikson, and Spyros Matsoukas

    Satyapriya Krishna, Andy Zou, Rahul Gupta, Eliot Krzysztof Jones, Nick Winter, Dan Hendrycks, J. Zico Kolter, Matt Fredrikson, and Spyros Matsoukas. D-REX: A Benchmark for Detecting Deceptive Reasoning in Large Language Models, September 2025. URL http://arxiv.org/abs/2509.17938. arXiv:2509.17938 [cs]

    arXiv 2025

  70. [71]

    Personal experiences bridge moral and political divides better than facts.Proceedings of the National Academy of Sciences, 118(6): e2008389118, February 2021

    Emily Kubin, Curtis Puryear, Chelsea Schein, and Kurt Gray. Personal experiences bridge moral and political divides better than facts.Proceedings of the National Academy of Sciences, 118(6): e2008389118, February 2021. ISSN 0027-8424, 1091-6490. doi: 10.1073/pnas.2008389118. URL https://pnas.org/doi/full/10.1073/pnas.2008389118

    work page doi:10.1073/pnas.2008389118 2021

  71. [72]

    König and Thomas Waldvogel

    Pascal D. König and Thomas Waldvogel. What matters for keeping or losing support in televised debates.European Journal of Communication, 37(3):312–329, June 2022. ISSN 0267-3231. doi: 10.1177/02673231211046706. URLhttps://doi.org/10.1177/02673231211046706

    work page doi:10.1177/02673231211046706 2022

  72. [73]

    Detecting Winning Arguments with Large Language Models and Persuasion Strategies, January 2026

    Tiziano Labruna, Arkadiusz Modzelewski, Giorgio Satta, and Giovanni Da San Martino. Detecting Winning Arguments with Large Language Models and Persuasion Strategies, January 2026. URL http://arxiv.org/abs/2601.10660. arXiv:2601.10660 [cs]

    arXiv 2026

  73. [74]

    White, Adam J

    Hause Lin, Gabriela Czarnek, Benjamin Lewis, Joshua P. White, Adam J. Berinsky, Thomas Costello, Gordon Pennycook, and David G. Rand. Persuading voters using human–artificial intelligence dialogues. Nature, pages 1–8, December 2025. ISSN 1476-4687. doi: 10.1038/s41586-025-09771-9. URL https://www.nature.com/articles/s41586-025-09771-9

    work page doi:10.1038/s41586-025-09771-9 2025

  74. [75]

    Wisniewski, Jin-Hee Cho, Sang Won Lee, Ruoxi Jia, and Lifu Huang

    Minqian Liu, Zhiyang Xu, Xinyi Zhang, Heajun An, Sarvech Qadir, Qi Zhang, Pamela J. Wisniewski, Jin-Hee Cho, Sang Won Lee, Ruoxi Jia, and Lifu Huang. LLM Can be a Dangerous Persuader: Empirical Study of Persuasion Safety in Large Language Models, April 2025. URL http://arxiv.org/abs/ 2504.10430. arXiv:2504.10430 [cs]

    arXiv 2025

  75. [76]

    Affective Interaction: Understanding, Evaluating, and Designing for Human Emotion.Reviews of Human Factors and Ergonomics, 7(1): 197–217, September 2011

    Danielle Lottridge, Mark Chignell, and Aleksandra Jovicic. Affective Interaction: Understanding, Evaluating, and Designing for Human Emotion.Reviews of Human Factors and Ergonomics, 7(1): 197–217, September 2011. ISSN 1557-234X. doi: 10.1177/1557234X11410385. URL https://doi. org/10.1177/1557234X11410385

    work page doi:10.1177/1557234x11410385 2011

  76. [77]

    Argument Strength is in the Eye of the Beholder: Audience Effects in Persuasion

    Stephanie Lukin, Pranav Anand, Marilyn Walker, and Steve Whittaker. Argument Strength is in the Eye of the Beholder: Audience Effects in Persuasion. In Mirella Lapata, Phil Blunsom, and Alexander Koller, editors,Proceedings of the 15th Conference of the European Chapter of the Association for Computational Linguistics: Volume 1, Long Papers, pages 742–753...

    2017

  77. [78]

    Enhancing LLM-Based Persuasion Simulations with Cultural and Speaker-Specific Information

    Weicheng Ma, Hefan Zhang, Shiyu Ji, Farnoosh Hashemi, Qichao Wang, Ivory Yang, Joice Chen, Juanwen Pan, Michael Macy, Saeed Hassanpour, and Soroush V osoughi. Enhancing LLM-Based Persuasion Simulations with Cultural and Speaker-Specific Information. In Christos Christodoulopoulos, Tanmoy Chakraborty, Carolyn Rose, and Violet Peng, editors,Findings of the ...

    work page doi:10.18653/v1/2025.findings-emnlp.808 2025

  78. [79]

    Reliability and Validity of Real-Time Response Measurement: a Comparison of Two Studies of a Televised Debate in Germany

    Jürgen Maier, Marcus Maurer, Carsten Reinemann, and Thorsten Faas. Reliability and Validity of Real-Time Response Measurement: a Comparison of Two Studies of a Televised Debate in Germany. International Journal of Public Opinion Research, 19(1):53–73, March 2007. ISSN 0954-2892. doi: 10.1093/ijpor/edl002. URLhttps://doi.org/10.1093/ijpor/edl002

    work page doi:10.1093/ijpor/edl002 2007

  79. [80]

    Chen, Dokyun Lee, and Michael D

    Emaad Manzoor, George H. Chen, Dokyun Lee, and Michael D. Smith. Influence via Ethos: On the Persuasive Power of Reputation in Deliberation Online.Management Science, 70(3):1613–1634, March

  80. [81]

    doi: 10.1287/mnsc.2023.4762

    ISSN 0025-1909, 1526-5501. doi: 10.1287/mnsc.2023.4762. URL https://pubsonline. informs.org/doi/10.1287/mnsc.2023.4762. 16

    work page doi:10.1287/mnsc.2023.4762 1909

Showing first 80 references.