# InquireMobile: Teaching VLM-based Mobile Agent to Request Human Assistance via Reinforcement Fine-Tuning Source: [InquireMobile: Teaching VLM-based Mobile Agent to Request Human Assistance via Reinforcement Fine-Tuning]() Collection index: [llms.txt]() Data updated: 2026-04-29T05:00:05.003Z Authors: Qihang Ai, Pi Bu, Yue Cao, Yingyao Wang, Jihao Gu, Jingxuan Xing, Zekun Zhu, Wei Jiang, Zhicheng Zheng, Jun Song, Yuning Jiang Published: 2025-08-27T08:40:05.000Z ## Abstract Recent advances in Vision-Language Models (VLMs) have enabled mobile agents to perceive and interact with real-world mobile environments based on human instructions. However, the current fully autonomous paradigm poses potential safety risks when model understanding or reasoning capabilities are insufficient. To address this challenge, we first introduce \\textbf{InquireBench}, a comprehensive benchmark specifically designed to evaluate mobile agents' capabilities in safe interaction and proactive inquiry with users, encompassing 5 categories and 22 sub-categories, where most existing VLM-based agents demonstrate near-zero performance. In this paper, we aim to develop an interactive system that actively seeks human confirmation at critical decision points. To achieve this, we propose \\textbf{InquireMobile}, a novel model inspired by reinforcement learning, featuring a two-stage training strategy and an interactive pre-action reasoning mechanism. Finally, our model achieves an 46.8% improvement in inquiry success rate and the best overall success rate among existing baselines on InquireBench. We will open-source all datasets, models, and evaluation codes to facilitate development in both academia and industry. ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [Inquire-driven GUI Navigation on InquireBench Chinese](): Inquire-driven GUI Navigation; InquireBench Chinese; SR: 10.5. - [Human Satisfaction Evaluation on InquireMobile EN](): Human Satisfaction Evaluation; InquireMobile EN; Satisfaction Score (0-5): 2.3. - [Human Satisfaction Evaluation on InquireMobile CN](): Human Satisfaction Evaluation; InquireMobile CN; Human Satisfaction Score: 2.2. - [Human Satisfaction Evaluation on InquireMobile Average](): Human Satisfaction Evaluation; InquireMobile Average; Human Satisfaction Score: 2.25. - [Generic GUI Execution on GUI-Odyssey](): Generic GUI Execution; GUI-Odyssey; Type Match (TM): 0.7237. - [GUI Execution on Android Control Low Instruction Abstraction](): GUI Execution; Android Control Low Instruction Abstraction; Type Match (TM): 91. - [GUI Execution on Android Control High Instruction Abstraction](): GUI Execution; Android Control High Instruction Abstraction; Type Match (TM): 71.9. - [Inquire-driven GUI Navigation on InquireBench English](): Inquire-driven GUI Navigation; InquireBench English; SR: 6.3. ## Original sources and implementation [Original paper]() [PDF]()