ResearchBenchmarksAgentic Mobile Interaction on AndroidWorld unseen tasks (test)Follow36.7Pass@1SGE15.921.326.732.1Mar 2, 2026Evaluation ResultsMethodMethodLinksPass@1SGERL Training=Strategy-G...RL Training=Strategy-Guided Exploration (SGE), Aggregation=Average ± standard error across 3 random seeds2026.0336.7GRPORL Training=GRPO, Aggr...RL Training=GRPO, Aggregation=Average ± standard error across 3 random seeds2026.0321.9Zero-ShotEvaluation Protocol=Di...Evaluation Protocol=Direct base-LLM evaluation2026.0316.7