微软发布Echoverse：深度演化环境提升计算机代理训练效果

AI 摘要

微软研究院发布了Echoverse，一个用于训练计算机使用代理的深度演化环境集合，包含十二个训练世界，强调高保真度而非数量。实验表明，在Echoverse上训练的9B模型得分从36.5%提升至67.1%，接近GPT-5.4。微软还开源了其中四个世界及其代码、数据和验证器，以促进该领域研究。

正文节选

Scaling fidelity over sheer count, targeting the capabilities agents actually lack, and evolving with the models they train. At a glance We built twelve training worlds for computer-use agents: ten deep domain worlds and two capability worlds, each drilling a single control rendered in many forms (date pickers and nested filters). Depth is what makes them worth training on: these worlds reproduce an application’s real behavior, come seeded with realistic data, and keep state coherent across scre

发布时间：2026-07-31 01:00

抓取时间：2026-08-02 00:26

来源机构：Microsoft Research