MemU-Bench: Benchmarking GUI Agents on Long-Horizon User Interactionswith Zhengxi Lu, Fei Tang, Yiwen Qiu, Yong Du +101 versionW_pdpqd5k9v1 · current