CS Paper Radar

Do LLMs Change Their Minds Like Humans? Diagnosing Human--LLM Divergence in Single-Turn Persuasion Judgments

Lin Chen, Yitong Chen, Yong Li

cs.CY cs.CL ·submitted 2026-08-30 14:08 UTC · v1
top venuecode/data

Abstract

Large language models (LLMs) are increasingly deployed as proxies for human participants in social simulations, yet whether they update their beliefs in response to persuasive arguments, as humans do, remains poorly understood. We conduct a systematic comparison using a naturally occurring online persuasion corpus in which original posters explicitly verify whether a reply changed their view. Our results show that LLMs achieve only slight agreement with humans (Cohen's kappa ranging from 0.079 to 0.178). Content-level analyses show that humans and LLMs agree on the strongest persuasion cues but diverge on finer ones: humans are more swayed by novel content and assertive language, whereas LLMs favor topical similarity and surface-level formatting. At the level of persuasion strategy, LLMs underweight emotional appeals and overweight credibility signals relative to humans, while the type of proposition under debate exerts no measurable effect on the degree of divergence. Furthermore, switching from first-person role-playing to third-person observation shifts all models toward greater resistance to persuasion, with the effect varying across persuasion strategies and textual features. These findings highlight the risk of treating LLM judgments as faithful proxies for human belief updating and point to structural differences in how LLMs and humans process persuasive discourse. Our code is available at https://github.com/tsinghua-fib-lab/LLM-belief-update-cmv.

Closest papers

Nearest neighbours by meaning across the whole index.

PerSo: Investigating Facets of LLMs' Social Intelligence in Real-World Persuasion Scenarios

cs.CL cs.AIcs.SI ·2026-10-14 ·Ali Zamani, Soroush Esfahanian, Erfan Shahabi, Mohammad Taher Pilehvar, Azadeh Shakery +1 more
benchmark similarity 0.76
Abstract · 25 pages, 17 figures, 10 tables. Ali Zamani and Soroush Esfahanian contributed equally

Understanding the social intelligence of large language models (LLMs) is increasingly important for applications involving human-AI interaction and multi-agent environments. Social intelligence comprises two aspects: cognitive intelligence, the ability to infer others' mental states, and behavioral intelligence, the ability to select appropriate social actions based on these inferences to achieve social goals. Persuasion provides a natural testbed in which both aspects are simultaneously at play, yet prior benchmarks typically evaluate them in isolation using synthetic scenarios. We introduce PerSo, a benchmark that evaluates both aspects on real-world persuasion dialogues from Reddit's ChangeMyView community, using an LLM-in-the-loop pipeline to construct the evaluation tasks, validated through human annotation. Our evaluation of diverse LLMs shows that performance generally scales with model size but varies across dimensions. The strongest LLM slightly surpasses the best observed human performance on the behavioral task, while the best human participant remains ahead on the cognitive task.

0.18

To Manipulate or Not: Opinion Dynamics in LLM Interactions

cs.AI cs.CLcs.MAcs.SI ·2026-10-20 ·Mohammad Siahkamari, Phuoc Dat Nguyen, Stephen MacNeil
similarity 0.74
Abstract

Large language models increasingly take part in extended conversations where they can both shape and be shaped by another party's views, yet most persuasion research measures only one side of that exchange. We introduce a controlled LLM-LLM debate framework in which two language-model roles repeatedly discuss a contested topic (climate-change attribution): one is given an assigned stance and, in some conditions, an explicit instruction to persuade, while the other receives no assigned position. Across four ordered model combinations involving GPT-6 Luna and Gemini 2.5 Flash Lite, we track both each role's explicitly reported position and a separate measure of its turn-by-turn behavior inside the conversation. Opinion dynamics differ substantially across model combinations: same-model interactions narrow the opinion gap on average, clearly for only one of the two models, while cross-model interactions widen it. Explicit reports and in-dialogue behavior capture distinct aspects of opinion dynamics, and a blanket instruction to persuade does not reliably increase influence. These findings suggest that conversational influence depends not only on assigned roles and initial positions but also on the participating models, with implications for how opinion change is measured in extended LLM interactions.

0.06

Evaluating the Capabilities of LLMs for Persuasive Dialogue

cs.CL ·2026-08-30 ·Jordan Robinson, Angus R. Williams, Katie Atkinson, Anthony G. Cohn
similarity 0.74
Abstract · 19 pages, 7 figures, 4 tables

Large language models (LLMs) can generate apparently highly persuasive text, but does sounding persuasive mean arguing well? We introduce \textsc{Persuasio}, a multi-agent dialogue platform grounded in a formal argumentation-based theory of persuasion dialogues that adjudicates logical winners during free-text debates. Using this system, we generated 192 debates on a UK political topic between humans and LLMs, and evaluated 22 interlocutors through both automated adjudication and 9,702 crowdsourced pairwise judgements across 1{,}386 annotation instances. We observed a consistent decoupling between subjective and formal persuasiveness: LLMs dominated the subjective ranking yet performed substantially worse under argumentation-theoretic adjudication, where humans remained competitive. Multi-agent and retrieval-augmented variants further widened this divergence. These findings reveal a systematic gap between rhetorical fluency and formal argumentative strength in LLM-based persuasive dialogues.

0.28

Belief Cascades Drive Persuasion in LLM Agent Networks

cs.CL cs.AI ·2026-08-25 ·Haoyi Qiu, Genglin Liu, Pranav Narayanan Venkit, Kung-Hsiang Huang, Saadia Gabriel +2 more
similarity 0.73
Abstract

Multi-agent LLM systems increasingly debate answers, coordinate research, simulate users, and mediate information flows, making agent-to-agent persuasion a basic but undermeasured capability. We introduce a controlled testbed for studying how goal-directed persuaders shift elicited stances in networks of LLM agents grounded in real-world ego-network topologies. Across four LLM backbones, five graphs, and 55 policy statements, we find that persuasion dynamics depend on the interaction between topology, competition, topic, and model prior. Additionally, we show that direct exposure reliably predicts next-round stance change in competing runs, and peer relays carry smaller but measurable influence, showing that agents not assigned to persuade can still transmit persuasive force. Finally, analyzing post text alone misses important movement: planned strategies are only partly realized in executed messages, action choices can diverge from message content, and persuadees rarely state the stance shifts detected by probes. These results argue for evaluating multi-agent persuasion as a trajectory- and exposure-level process, using belief probes, exposure provenance, and action logs to identify who influenced whom and whether visible language reflects underlying stance movement.

0.34

A Thousand Faces: A Persuasion Benchmark with Experience-Grounded Persuadees Informed by Psychology and Pragmatics

cs.CL ·2026-10-18 ·Ziqing Zhuang, Dingyi Zhang, Deyu Zhou
benchmarkcode/data similarity 0.73
Abstract · 45 pages. Dataset: https://huggingface.co/datasets/Slandes/ATF-Persu

People can express the same beliefs and desires yet need different reasons to change their minds. What makes an argument relevant depends on how those views arose: experiences shape what a person finds credible, what matters to them, and what could prompt reconsideration. These personal grounds may also be expressed indirectly, through contextual clues in conversation. Benchmarks that specify beliefs and desires without these connections can reward arguments that match a description while overlooking whether they address the person's reasons. We introduce A Thousand Faces (ATF-Persu), with 1,000 personas across 200 topics and nine domains. Drawing on personality and attitude theories and conversational implicature, we construct five personas per factual context, linking experiences to judgments, expression, and grounds for revision. Evaluations show lower persuasion success than three comparison benchmarks under their respective protocols. Complete backgrounds yield broader use of supplied personal grounds and higher persona faithfulness than simplified specifications; blinded reviewers also rate ATF-Persu above CToMPersu on identifiable personal reasons, their influence on responses, and contextual expression. The dataset is publicly available at https://huggingface.co/datasets/Slandes/ATF-Persu.

0.31

User Attributes and Political Ideology Shape LLM Receptiveness in Online Political Debates

cs.AI cs.CL ·2026-11-02 ·Akram Elbouanani, Yousra Taleb, Gianluca Demartini, Adrian Popescu
similarity 0.72
Abstract · Preprint. Submitted for conference review

Online political discussions increasingly involve AI-mediated participation, as users invoke LLMs to assess contested claims, arbitrate disputes, and respond to opposing viewpoints. As these systems become participants in online exchanges, they may alter which speakers and argumentative strategies receive validation within the discussion. We study this interactional dimension by reconstructing 5,376 Debate.org debates as LLM conversations and running nearly 150,000 controlled multi-turn conversations that vary user political orientation, expertise, rhetorical strategy, and defended position, including conversations with competing users. Across naturalistic and controlled debates, we find that LLMs are systematically more receptive to left-leaning than right-leaning users. The asymmetry persists when users with different political orientations defend the same position and when opposing users appear in the same conversation, indicating that it is not fully explained by stance agreement or simple accommodation to the current speaker. Repeated interaction increases receptiveness without eliminating this ordering; expertise also matters, while rhetorical differences are smaller. The pattern appears across multiple models, but is not fixed: fine-tuning on different political corpora can largely erase or reverse it. These results show that LLMs can respond systematically differently to participants in the same political debate, making such interactional asymmetries important for understanding their emerging role in online public discourse.

0.20

Does Personality Persuade? Investigating Personality Trait Prompting and Persuasion in Large Language Models

cs.CY ·2026-11-04 ·Ruben Boone, Gilles Eerlings, Sebe Vanbrabant, Davy Vanacken, Kris Luyten
similarity 0.72
Abstract

Large language models (LLMs) can be prompted to express different personality profiles, but it remains unclear how reliably they do so or whether personality affects their persuasiveness. We investigate these questions in two evaluations. First, we assess Big Five personality prompting using the IPIP-NEO-120, examining reproducibility, trait controllability, cross-trait effects, and unconditioned model profiles. Second, we conduct a within-subjects user study (N=32) in which participants discuss moral dilemmas with LLM agents whose personalities are aligned, mismatched, neutral, or unconditioned. We find that prompted profiles can be reproducible and responsive to requested traits, although controllability varies across models and changes to one trait can affect others. Unconditioned models also exhibit systematic personality profiles rather than neutral baselines. In the user study, personality alignment did not significantly affect perceived persuasiveness, stance switching, or confidence changes. Participants switched their initial stance in only 8 of 128 conversations, while confidence changes were more common. Participants' Big Five traits were not significantly associated with susceptibility to persuasion, but higher self-reported AI persuasion literacy was associated with less frequent confidence changes. Exploratory analysis showed that logical appeals dominated agent messages. These findings distinguish technical control over prompted LLM personality from observable influence on users and motivate evaluating persuasive interactions beyond explicit changes in decisions.

0.16

Learning to Persuade Exposes How Easily LLMs Abandon Correct Beliefs

cs.CL cs.AI ·2026-08-12 ·Nimet Beyza Bozdag, Emre Can Acikgoz, Gokhan Tur, Dilek Hakkani-Tür
top venue similarity 0.71
Abstract · To be published in NeurIPS 2026

Persuasion is a core dynamic of natural language communication, shaping how large language models (LLMs) update beliefs, resolve disagreements, and reach decisions. As LLMs increasingly debate, advise, and think collaboratively with humans and each other, resistance to harmful persuasion becomes a core requirement for reliable behavior. Yet a single targeted persuasive argument can be enough to collapse model accuracy to near zero, even when the argument is factually false. We formalize this threat as adversarial persuasion and introduce an adversarial reinforcement learning framework that trains persuader agents to change a target model's answer in a single interaction. First, we show that RL-trained persuaders raise persuasion success from approximately 24% to over 93% against the training-time persuadee. Second, we find that these learned strategies generalize across four out-of-distribution benchmarks, maintaining 89-95% persuasion success against the training-time persuadee. They also transfer to unseen models, achieving 83% success on Qwen-14B, 79% on Llama-3.1-8B, and 25% on GPT-4o-mini, although effectiveness can be limited for frontier models, with only 4% success on GPT-5-mini. Third, we demonstrate that a curriculum that bootstraps on more persuadable open-weight models before targeting harder models further increases GPT-4o-mini attack success from 25% to 38%. Moreover, our results reveal that optimized persuaders increasingly rely on credibility-based tactics, including fabricated citations and false authoritative evidence. Together, these findings expose a critical weakness in current LLM agents. Even when they initially reason correctly, they can be steered toward false conclusions by optimized natural language influence. This positions persuasion robustness as a necessary safety criterion for multi-agent and human-AI decision-making systems.

0.66