AI RESEARCH
MobiFlow: Real-World Mobile Agent Benchmarking through Trajectory Fusion
arXiv CS.AI
•
ArXi:2604.09587v1 Announce Type: new Mobile agents can autonomously complete user-assigned tasks through GUI interactions. However, existing mainstream evaluation benchmarks, such as AndroidWorld, operate by connecting to a system-level Android emulator and provide evaluation signals based on the state of system resources.