让AI在推理游戏中学会说服他人，提升社交影响力。
The Stackelberg Speaker: Optimizing Persuasive Communication in Social Deduction Games
- 将对话建模为领导者-追随者博弈，策略性影响对手决策
- 在3个不同推理游戏上表现超越基线，显著提升说服效果
- 适合研究社交影响、人机协作或博弈策略的学者
大型语言模型（LLM）代理在社会推理游戏（SDGs）中展现出显著进展。然而，现有方法主要关注信息处理与策略选择，忽视了说服性沟通在影响其他玩家信念与反应中的关键作用。在SDGs中，成功不仅取决于正确推理，更在于说服他人按自身意图行动。为此，我们将回合制对话形式化为斯塔克尔伯格博弈，当前玩家作为领导者，战略性地影响跟随者的回应。基于这一理论框架，我们提出一种强化学习方法，训练代理优化话语以实现最大说服力。在三个不同社会推理游戏上的全面实验表明，我们的代理显著优于基线模型。该工作标志着向具备战略社交影响力的人工智能代理迈出重要一步，对需说服沟通的场景具有广泛意义。代码与数据见 https://3dagentworld.github.io/leader_follower。
原文摘要 · Abstract (English)
Large language model (LLM) agents have shown remarkable progress in social deduction games (SDGs). However, existing approaches primarily focus on information processing and strategy selection, overlooking the significance of persuasive communication in influencing other players' beliefs and responses. In SDGs, success depends not only on making correct deductions but on convincing others to response in alignment with one's intent. To address this limitation, we formalize turn-based dialogue in SDGs as a Stackelberg competition, where the current player acts as the leader who strategically influences the follower's response. Building on this theoretical foundation, we propose a reinforcement learning framework that trains agents to optimize utterances for persuasive impact. Through comprehensive experiments across three diverse SDGs, we demonstrate that our agents significantly outperform baselines. This work represents a significant step toward developing AI agents capable of strategic social influence, with implications extending to scenarios requiring persuasive communication. Our code and data are available at https://3dagentworld.github.io/leader_follower.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。