Jul 2026
SCALECUA: Scaling Computer Use Agents with Verifiable Task Synthesis and Efficient Online RL
This work introduces ScaleCUA, a unified framework that scales online RL for CUAs via verifiable task synthesis and efficient training, and designs VeriGen, an end-to-end framework for generating verifiable RL tasks through iterative docker interactions and a multi-agent feedback loop.
Bowen Lv, Xiao Liu, Yanyu Ren et al.
· arXiv.org · 4 citations
· ⚡1