微软 Agensh 扩到 1024 个 Agent,pandoc 通过率升至 55%
MicrosoftResearch · hf · 2026-09-23
Microsoft Research introduced Agensh, a self-organized multi-agent harness with no central orchestrator, exploring agent count as a new scaling dimension.
- Motivation: existing multi-agent harnesses are often bottlenecked by a central orchestrator's capacity to allocate tasks. Agensh instead runs a cooperation loop where concurrent workers gather context, self-assign subtasks, act, share findings, verify, and merge progress asynchronously.
- Infrastructure: three components—a shared workspace for proposed/ongoing/completed work, a message interface for worker communication, and shared context retaining reusable findings and intentions.
- Results: on the five hardest ProgramBench tasks with GPT-5.6-sol (high), scaling from 1 to 128 agents raised mean test-pass rate from 19.31% to 28.78% (49% relative improvement); larger organizations reach comparable rates earlier.
- On pandoc, scaling from 1 to 1,024 agents lifted the pass rate from 33.89% to 55.06%. Worker trajectories show self-organized cooperation forms emerging and standardizing as the organization grows.
所属事件:微软 Agensh：无中心编排扩到 1024 个 Agent(3 条相关)→
「编程与Agent」频道最新
- Plurai 推出 Agent 仿真平台：从代码库构建数字孪生自动找故障 — thisguyknowsai · 2026-11-11
- 0xJeff:x402 的真正短板在企业级,大钱在 B2V 而非机器互付 — 0xJeff · 2026-11-11
- 开发者自建 git 网页应用被 Codex 安全护栏误判为网络安全风险 — brianmichel · 2026-11-11
- 形式化验证让 AI 编码更快：Lean 证明成为 agent 优化护栏 — bigdata · 2026-11-11
- Orchestrator v0.14.11 发布：统一管理散落各处的 AI 编程会话 — julianweisser · 2026-11-11
- DeepSeek 开源编码 agent DSH 登上 HF Harness 榜，五项评测并列第一 — TheZachMueller · 2026-11-11