Recovering Policy-Induced Errors: Benchmarking and Trajectory Synthesis for Robust GUI Agents
About
While GUI agents have advanced rapidly, they often lack the robustness to recover from their own errors, hindering real-world deployment. To bridge this gap at both the evaluation and data levels, we introduce GUI-RobustEval and propose Robustness-driven Trajectory Synthesis. GUI-RobustEval contains $1,216$ executable test cases that systematically measure error recovery capabilities across a broad and realistic spectrum of error modes. At the data level, RoTS is a scalable synthesis framework that creates $800k$ high-quality data via a tree-based pipeline that proactively discovers diverse error modes and synthesizes corresponding recovery steps. Our two models, RoTS-7B and RoTS-32B, fine-tuned on our dataset, both demonstrate significant gains on GUI-RobustEval and traditional GUI benchmarks. Notably, RoTS-32B achieves state-of-the-art performance on OSWorld, with a $47.4\%$ success rate and a $33.8\%$ All-Pass@4 score, suggesting that improved long-horizon error recovery ability contributes to both robustness and overall performance. Our code is available at https://github.com/AlibabaResearch/RoTS.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Windows UI Navigation | WindowsAgentArena (WAA) | Success Rate39.1 | 33 | |
| GUI Agent Interaction | OSWorld | Success Rate (Max Steps: 15)42.8 | 16 | |
| Robust GUI Navigation | GUI-RobustEval | Success Rate (Depth 0)49.7 | 12 |