论文

研究:多个 AI 智能体共享资源利用率不足

Groups of AI agents underuse resources despite shared evidence of increased capacity

精选理由

多智能体协调的老问题有了新实验证据:光告诉 Claude Haiku 5.5、GPT-6.1 Sol 这些模型"服务器可能恢复了"没用,给它们看一条成功记录才管用。做 Agent 系统负载均衡的人值得拿来对齐自己的信息共享设计。

arXiv 论文(arXiv:2611.22632)测试了语言模型智能体在共享有限资源时的协调行为:智能体在两个模拟服务器之间分配任务,其中一台已恢复但状态报告仍标注不可用。实验显示,仅提示"恢复是可能的"能减少错过截止日期的任务数;在 Anthropic Claude Haiku 5.5 的对照测试中,收到规则提醒后没有任何智能体尝试已恢复的服务器,但看到一条成功请求记录后全部选择了该服务器。

正文

团队进一步保持成功记录不变,向智能体分发数字信号:使用 OpenAI GPT-6.1 Sol 或 Google Gemini 3.5 Flash-Lite 时,给组内成员分配不同数字能让并发请求分散到不同服务器,任务完成数上升,而整体选择频率几乎不变。不过信息共享后,测试的全部五个模型仍会留下未使用的容量。

结论指向两个设计挑战:告知智能体资源可用性只是第一步,还需要机制帮助它们把需求分散到各资源上。这对多智能体系统开发者在负载均衡和信息共享机制设计上有直接参考价值。

原文 · arXiv: Anthropic

Groups of AI agents underuse resources despite shared evidence of increased capacity

When AI agents share limited resources, such as computing services or public infrastructure, their individual choices affect how much work they complete as a group. Can sharing evidence of what is available let a group make full use of its resources after capacity changes? We tested this with language-model agents choosing between two simulated servers. One server recovered while its status report still listed it as unavailable. Each server had limited capacity per round, so rejected jobs required another attempt before the deadline. Telling agents that recovery was possible reduced missed deadlines. In controlled tests with Anthropic's Claude Haiku 5.5, no agent tried the recovered server after a rule reminder, yet all chose it when shown a successful request record. After sharing, all five models tested left capacity unused. Keeping the success record fixed, we gave agents numbers without instructions, using each value equally often but giving group members either identical or distinct numbers. With OpenAI's GPT-6.1 Sol or Google's Gemini 3.5 Flash-Lite, distinct numbers spread simultaneous requests across servers and increased job completions even though overall choice frequencies barely changed. Telling AI agents which resources are available leaves a second design challenge: helping them spread demand across those resources.

产业坐标与信源核验

本条情报归属 AITOP「论文」核心栏目,重点聚焦北美前沿实验室与硅谷科技巨头。在跨区域产业观察坐标中,该动态直接关联「AI 开发者工具链、Coding Agent 与工程化实践」研究主线,反映出当前技术演进与商业落地的最新风向。涉及核心实体与技术议题包括:多智能体、arXiv、Claude、GPT-6.1 Sol,读者可通过下方标签进入对应专题页查阅历史脉络与关联对比。

信源核验说明:本文原始信息发布自一手权威信源「arXiv: Anthropic」(Tier-1 一手发布),具备高确定性参考价值。