arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2605.09128cs.MAcs.AI
收藏

内部 vs. 外部:比较 deliberation 与 evolution 在多智能体宪法设计中的应用

Internal vs. External: Comparing Deliberation and Evolution for Multi-Agent Constitutional Design

College of Engineering, Shibaura Institute of Technology, Tokyo, Japan(东京工业大学工学院);Department of EECS, University of California, Berkeley, Berkeley, CA, USA(加州大学伯克利分校电子工程与计算机科学系)

Hershraj Niranjani, Ujwal Kumar, Phan Xuan Tan

AI总结:

本文比较了多智能体系统中内部协商与外部进化在协调网格世界、公共物品游戏和双边交易市场中的表现,发现进化在集体行动中表现更优,但在双边交易中无明显优势,且进化在激励变化时可能产生负面效果。

摘要:

多智能体AI系统需要行为宪法,但尚不清楚这些规则应通过智能体自我治理内部产生还是通过优化外部发现。本文首次在三个社会环境中对内部协商和外部进化进行了受控比较:协调网格世界、重复公共物品游戏和双边交易市场。在180次模拟运行中,进化在集体行动场景中显著优于协商(p < 0.01),而在双边交易中两种方法均无提升。乘数消融显示,进化在激励变化时优势消失:当池乘数(m = 0.75)时,进化宪法强制价值毁灭型合作,成为最差方法。值得注意的是,三十次试验中无任何协商运行提出惩罚——这是进化可靠发现的维持合作机制,表明外部优化在峰点胜出,而内部自治理则以峰点换取结构响应性。

英文摘要:

Multi-agent AI systems need behavioral constitutions, but it is unresolved whether such rules should emerge internally through agent self-governance or be discovered externally through optimization. We present the first controlled comparison of internal deliberation and external evolution across three social environments: a coordination grid-world, an iterated public goods game, and a bilateral trading market. Across 180 simulation runs, evolution significantly outperforms deliberation in collective-action settings (p < 0.01), while neither method improves outcomes in bilateral trading. A multiplier ablation reveals that evolution's advantage inverts when incentives shift: at pool multiplier (m = 0.75) the evolved constitution forces value-destroying cooperation and becomes the worst-performing method. Notably, no deliberation run across thirty trials ever proposed punishment -- the canonical cooperation-sustaining mechanism evolution reliably discovers -- suggesting external optimization wins on peaks while internal self-governance trades peaks for structural responsiveness.

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。

↑