发表机构
Microsoft Research（微软研究院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
Agensh提出无中央协调器的自组织多智能体框架，通过并发协作循环将智能体扩展至1,024个，显著提升任务通过率，开辟智能体数量作为新的扩展维度。
AI 中文摘要
多智能体系统可以通过并发执行工作来降低复杂任务的延迟。几个开创性的框架（harness）支持多智能体系统。然而，当前多智能体框架（harness）的可扩展性往往受到中央协调器分配任务和协调工作能力的限制。为了解决这一局限，我们引入了Agensh，一种无需中央协调器的可扩展自组织多智能体框架（harness）：并发工作智能体执行多智能体协作循环，持续收集上下文，认领并自行分配子任务，采取行动并分享发现，验证结果，并以异步方式合并进展。该循环由智能体组织基础设施支持，该基础设施包含三个组件：共享工作区保存提议中、进行中和已完成的工作；消息接口让智能体进行通信；共享上下文保留可复用的发现和工作意图。为了测试Agensh的可扩展性，我们使用GPT-5.6-sol（高）在五个最难的ProgramBench任务上对其进行评估。将智能体数量从1扩展到128个，平均最终测试通过率从19.31%提高到28.78%，相对提升约49%。更大的组织更早达到相当的测试通过率。在pandoc任务上，将智能体数量从1扩展到1,024个，最终测试通过率从33.89%提高到55.06%。智能体的轨迹进一步表明，随着组织的壮大，不同形式的自组织协作逐渐涌现并标准化。这些结果揭示了智能体数量作为多智能体组织扩展通用智能前沿的一个新扩展维度，为在硬延迟约束或时间预算下的复杂任务提供了实用解决方案。
英文摘要
A multi-agent system can reduce latency on complex tasks by executing work concurrently. Several pioneering harness frameworks support multi-agent systems. However, the scalability of current multi-agent harnesses is often constrained by a central orchestrator's capacity to allocate tasks and coordinate workers. To address this limitation, we introduce Agensh, a scalable self-organized multi-agent harness without a central orchestrator: concurrent workers execute a multi-agent cooperation loop, continuously gathering context, claiming and self-assigning sub-tasks, taking action and sharing findings, verifying results, and merging progress in an asynchronous manner. The loop is supported by the agentic organization infrastructure comprising three components: a shared workspace holds proposed, ongoing, and completed work; a message interface lets workers communicate; and shared context retains reusable findings and work intentions. To test the scalability of Agensh, we evaluate it on the five hardest ProgramBench tasks with GPT-5.6-sol (high). Scaling from 1 to 128 agents raises the mean final test-pass rate from 19.31% to 28.78%, an approximately 49% relative improvement. Larger organizations reach comparable test-pass rates earlier. On pandoc, scaling from 1 to 1,024 agents raises the final test-pass rate from 33.89% to 55.06%. Worker trajectories further show that different forms of self-organized cooperation gradually emerges and standardizes as the organization grows. These results reveal the number of agents as a new scaling dimension for multi-agent organizations to expand the frontier of general intelligence, offering a practical solution for complex tasks under hard latency constraints or time budgets.
Comments13 pages, 6 figures
