[2603.24533] UI-Voyager: A Self-Evolving GUI Agent Learning via Failed Experience

Computer Science > Machine Learning arXiv:2603.24533 (cs) [Submitted on 25 Mar 2026] Title:UI-Voyager: A Self-Evolving GUI Agent Learning via Failed Experience Authors:Zichuan Lin, Feiyu Liu, Yijun Yang, Jiafei Lyu, Yiming Gao, Yicheng Liu, Zhicong Lu, Yangbin Yu, Mingyu Yang, Junyou Li, Deheng Ye, Jie Jiang View a PDF of the paper titled UI-Voyager: A Self-Evolving GUI Agent Learning via Failed Experience, by Zichuan Lin and 11 other authors View PDF HTML (experimental) Abstract:Autonomous mobile GUI agents have attracted increasing attention along with the advancement of Multimodal Large Language Models (MLLMs). However, existing methods still suffer from inefficient learning from failed trajectories and ambiguous credit assignment under sparse rewards for long-horizon GUI tasks. To that end, we propose UI-Voyager, a novel two-stage self-evolving mobile GUI agent. In the first stage, we employ Rejection Fine-Tuning (RFT), which enables the continuous co-evolution of data and models in a fully autonomous loop. The second stage introduces Group Relative Self-Distillation (GRSD), which identifies critical fork points in group rollouts and constructs dense step-level supervision from successful trajectories to correct failed ones. Extensive experiments on AndroidWorld show that our 4B model achieves an 81.0% Pass@1 success rate, outperforming numerous recent baselines and exceeding human-level performance. Ablation and case studies further verify the effectiveness of GRSD. O...