arXiv cs.MA· Gregory B. Rehm·· 9 天前AI 评分44
GovSim 研究：LLM 智能体能否在崩溃中自我治理并牺牲自己？
Can Agents Adapt Their Institutions And Sacrifice Themselves During Collapse? Executable Self-Governance in a GovSim Commons
AI 导读
研究者构建 GovSim-SelfGovern 公共资源实验，让 LLM 智能体用可执行 Python 立法、在沙盒中测试并投票。渔业资源充足时，自治将社区无人挨饿的存活概率从 52.5% 提升至 75.0%；但资源不足时，治理无法挽救共同体。投票设计影响巨大：沙盒预览立即点名被驱逐者时放逐法通过率仅 8%，推迟到未来危机时升至 55%；一句框架表述可将自我牺牲从 2.5% 拉至 91%。
正文 · 原文
Abstract:LLM agents are beginning to write the rules they live under: policies, contracts and code that other agents review and approve. What happens when those rules must decide who is sacrificed from a group? We built GovSim-SelfGovern, a commons in which agents legislate in executable Python, test each law in a sandbox, vote on it, and live with the result. When the fishery is abundant, self-rule helps, raising the chance that the community survives without anyone starving from 52.5% to 75.0%. When it cannot feed everyone though, adding governance is not enough to save the commons. We trace much of this to the vote: agents readily draft laws that remove members, but then frequently vote them down, thereby dooming the commons. In a controlled replay, the same exile law passes 8% of the time when the sandbox preview names its victim at once and 55% when the removal is deferred until a future crisis. Showing voters the full source code does not close the gap, though most describe the removal in their own words. Nor is sacrifice a fixed trait. A single sentence of framing can swing self-removal from 2.5% to 91%, and due process deters agents more than considering the potential death of a peer. In contrast, give the community a treasury and the question of who must go almost disappears. Our work shows that what our collectives legislate is decided more by the form in which a choice is put to them as compared to their ethical beliefs.
| Subjects: | Multiagent Systems (cs.MA) |
| Cite as: | arXiv:2609.22600 [cs.MA] |
| (or arXiv:2609.22600v2 [cs.MA] for this version) | |
| https://doi.org/10.48550/arXiv.2609.22600 arXiv-issued DOI via DataCite |
Submission history
From: Gregory Rehm [view email]
[v1]
Fri, 18 Sep 2026 21:36:30 UTC (1,750 KB)
[v2]
Wed, 7 Oct 2026 20:51:36 UTC (266 KB)
来源：arXiv cs.MA · arxiv.org