| license: apache-2.0 | |
| pipeline_tag: image-text-to-text | |
| library_name: transformers | |
| # UI-S1: Advancing GUI Automation via Semi-online Reinforcement Learning | |
| This repository contains the efficient GUI grounding model, **UI-S1-7B**, presented in the paper [UI-S1: Advancing GUI Automation via Semi-online Reinforcement Learning](https://huggingface.co/papers/2509.11543). | |
| Project page / Code: [https://github.com/X-PLUG/MobileAgent/tree/main/UI-S1](https://github.com/X-PLUG/MobileAgent/tree/main/UI-S1) | |
| ## Paper Abstract | |
| Graphical User Interface (GUI) agents have demonstrated remarkable progress in automating complex user interface interactions through reinforcement learning. However, current approaches face a fundamental dilemma: offline RL enables stable training on pre-collected trajectories, but struggles with multi-step task execution for lack of trajectory-level reward signals; online RL captures these signals through environment interaction, but suffers from sparse rewards and prohibitive deployment costs. To address it, we present Semi-online Reinforcement Learning, a novel paradigm that simulates online RL on offline trajectories. During each rollout process, we preserve the original model output within the multi-turn dialogue, where a Patch Module adaptively recovers the divergence between rollout and expert trajectories. To capture long-term training signals, Semi-online RL introduces discounted future returns into the reward computation and optimizes the policy with weighted step-level and episode-level advantages. We further introduce Semi-Online Performance (SOP), a metric that aligns better with true online performance, serving as a practical and effective proxy for real-world evaluation. Experiments show that ours Semi-online RL achieves SOTA performance among 7B models across four dynamic benchmarks, with significant gains over the base model (e.g., +12.0% on AndroidWorld, +23.8% on AITW), demonstrating significant progress in bridging the gap between offline training efficiency and online multi-turn reasoning. | |
| ## Overview | |
| We present **Semi-online RL**, a novel paradigm that simulates online reinforcement learning using offline trajectories, thereby enabling the efficient training of MLLM-based GUI agents with enhanced multi-turn interaction capabilities. | |
| <div align="center"> | |
| <img src="https://github.com/X-PLUG/MobileAgent/raw/main/UI-S1/assets/method_comparison.png" alt="Method Comparison" style="width:80%;"> | |
| </div> | |
| Ours **UI-S1-7B** achieves SOTA performance on both semi-online metric (SOP) and online metric (AndroidWorld) among open-source 7B models. | |
| <div align="center"> | |
| <img src="https://github.com/X-PLUG/MobileAgent/raw/main/UI-S1/assets/metric.png" alt="Metrics" style="width:80%;"> | |
| </div> | |
| ## Detailed results | |
| <div align="center"> | |
| <img src="https://github.com/X-PLUG/MobileAgent/raw/main/UI-S1/assets/result.png" alt="Results" style="width:80%;"> | |
| </div> |