使用多LLM智能体模拟仇恨言论级联：实证基础、建模保真度与干预策略
Simulating Hate Speech Cascades with Multi-LLM Agents: Empirical Grounding, Modeling Fidelity, and Intervention Strategies
AI总结：
本研究通过多LLM智能体系统模拟在线仇恨言论传播，发现其能再现实证数据中的立场单一性和毒性同质性，并通过消融实验识别出智能体异质性为关键保真因素，提出针对密集网络的放大器干预策略。
摘要：
在线平台上仇恨内容传播的忠实建模仍然是内容审核研究中的一个开放问题。经典的级联模型没有明确表示与仇恨内容传播相关的用户画像、社区和内容因素，因此在实际场景中部署时可能产生效果较差的审核策略。多智能体大语言模型系统原则上可以使每次转发决策依赖于用户画像、周围社区和帖子内容，但尚不清楚这种增加的灵活性是否比经典基线更忠实地再现真实的仇恨级联。我们研究了三个仇恨Bluesky级联和一个大小匹配的良性对照。在实证Bluesky数据中，我们发现：97.4--99.7%的转发者采取敌对立场；对于仇恨级联，扩散树上的毒性-参与同质性高于关注图；仇恨级联的拓扑结构是星形（大多数转发直接来自根节点），而良性级联是树形（转发通过多跳链传播）。在模拟中，多LLM智能体模拟器再现了立场单一性和毒性差异方向。结构化消融实验将智能体异质性识别为主要的保真因素，针对密集网络的放大器干预在5.7%良性附带损害下实现了7.5--12.9%的减少。
英文摘要：
Faithful modeling of hateful content propagation on online platforms remains an open problem for moderation research. Classical cascade models that do not explicitly represent the profile, community, and content factors associated with hateful-content propagation may yield moderation strategies that behave less effectively when deployed in real-world scenarios. Multi-agent large language model (LLM) systems can, in principle, make each reshare decision depend on the user's profile, the surrounding community, and the post's content, but it remains unclear whether this added flexibility actually reproduces real hateful cascades more faithfully than classical baselines. We study three hateful Bluesky cascades and a size-matched benign control. In the empirical Bluesky data, we found that: 97.4--99.7\% of reposters take a hostile stance; toxicity-engagement homophily is higher on the diffusion tree than on the follower graph for hateful cascades; topology is star-like for the hateful cascades (most reposts come directly from the root) versus tree-like for the benign cascade (reposts propagate through multi-hop chains). In simulation, a multi-LLM-agent simulator reproduces the stance monoculture and the toxicity-delta direction. A structured ablation identifies agent heterogeneity as the leading fidelity factor, and amplifier targeting on dense networks yields 7.5--12.9\% reduction at 5.7\% benign collateral.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
