Zhoujia Zhang

OS-MAP: How Far Can Computer-Using Agents Go in Breadth and Depth?

Learning to Think in Blocks: A Prior-Guided Reinforcement Learning Framework for RAG