Papers & Research

Microsoft Research releases Echoverse: a deeply evolvable training environment for computer-use agents

Original title:Echoverse: Deep, evolving environments for computer-use agents

AI summary

Microsoft Research introduced Echoverse, which builds 12 synthetic training worlds for computer-use agents (10 deep domain worlds and 2 capability worlds). After training, a 9B model's average score across all environments rose from 36.5% to 67.1%, 14 points behind GPT-5.4 (80.7%).

Key Takeaways

  • Microsoft Research introduced Echoverse,which builds 12 synthetic training worlds for computer-use…
  • After training,a 9B model's average score across all environments rose from 36.
  • 1%,14 points behind GPT-5.

Quick Facts

Published
Updated
Source
Microsoft Research 博客(RSS)
Category
Papers & Research
AI score
60
Read the original Read the full article at the original source
Why it matters

原文给出深度世界对比浅层世界的实验数据和可复用的环境构建方法,并开源了四个世界,读者可据此评估训练环境的做法。

Related entities

Source & copyright

This article is based on curated titles and AI summaries from the AIHOT public API, with attribution and links back. Original copyright remains with the original publishers and authors; this site does not mirror full texts.

Original article · AIHOT entry

How to Cite

When citing this article, keep the title, date, and permalink, and credit Orange News as the source.

Orange News: Microsoft Research releases Echoverse: a deeply evolvable training environment for computer-use agents. July 31, 2026. https://new.nat6.net/en/news/cmtyn6l7k0007rorjmk2pg8ui/

View all →