People often cooperate out of concern for their reputation. The corresponding theory of indirect reciprocity predicts that such cooperation can only evolve if people's opinions of each other are sufficiently correlated. This correlation, however, can be difficult to achieve when individuals form their opinions independently from one another, as in private assessment models. In that case, errors can generate disagreements that propagate throughout a population. Here, we identify a simple mechanism that mitigates this problem. Most prior work assumes that observers revise only the reputations of donors-the individuals who choose whether to cooperate or defect. We instead allow observers to also revise the reputations of recipients-the individuals affected by the donors' choices. Using analytical calculations and simulations, we show that such dual reputation updates help synchronize opinions and facilitate cooperation. To demonstrate this approach, we focus on a particularly simple rule, which we call Recipient Image Scoring (RIS). Under RIS, recipients are assessed as good whenever donors choose to cooperate with them. We show that this rule for judging recipients effectively complements the classical leading-eight rules for judging donors. Our results establish dual assessment as a simple mechanism for cooperation that relies only on directly observable information.
Yukari Jessica Tham, C. Hilbe, Yohsuke Murase· Proceedings of the National...· 0 citations
LLMs are increasingly deployed as proxies for human study participants in social science experiments, yet the fidelity of this practice has rarely been tested directly. We test whether six LLMs can simulate individual human belief updates, comparing LLM outputs 1-to-1 against ground truth data from 391 UK participants on Prolific, who updated their stances on three discussion topics after reading Reddit comments. Each participant was simulated by an LLM conditioned on a persona derived from their demographic and personality trait data. We find that some LLMs (Qwen3-32B and GPT-5-Mini) can match the human post-stance distribution, but only when given participants'actual initial stances. All six models fail to simulate initial stances themselves and to produce faithful belief updates from self-generated stances. Three systematic biases emerge across all models: overrepresentation of neutral positions, more frequent but smaller belief shifts than humans, and a failure to rank comments by convincingness. Demographic and personality trait personas had no consistent effect on fidelity. LLM simulations of human belief dynamics are only reliable when grounded in realistic starting conditions, that current multi-round social media simulations rarely provide.
Sebastian Pohl, Harsh Mehta, Pranav Mambayil et al.· arXiv.org· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
