Princeton 论文：多开发者专精智能体组合在 GAIA 上达 70%
Princeton 这篇论文挺有意思：几个开发者各自做的专精智能体拼起来，GAIA 分数翻倍到 70%，还告诉你什么时候组合会翻车。
Princeton 论文模拟了多个开发者各自构建 LLM 智能体的场景，比较单人系统与多开发者系统的表现。在 GAIA 基准上，同样开发预算下，单独最高仅 35% 得分的专精智能体组合后达到 70%，比通用智能体的中位数高 12 个百分点。在仅限网页搜索的 SealQA 上，组合系统表现反而更差且成本更高。一个单独得分仅 19% 的智能体能把组合系统从 40% 提升到 53%，作者据此建议开发者针对共享系统的空白点构建智能体。
New Princeton paper finds that combining specialized agents from many developers can beat any single developer's general agent, with GAIA accuracy reaching 70%.
Today most agentic systems are built and controlled by 1 developer, so each team rebuilds the same basic skills. The authors argue this limits scale.
They simulated populations of developers with LLM agents that build agents under a fixed budget, then compared lone systems with a multi-developer system. The multi-developer system splits each task and routes the parts to different developers' agents.
On GAIA with the same total development budget, specialists that each scored at most 35% alone reached 70% together. That is 12 points above the median general agent, while on web-search-only SealQA the combined system did worse at higher cost.
Build agents to fill gaps in shared systems, since an agent scoring 19% alone lifted a combined system from 40% to 53%.
