Research map / Image generation / Text-to-image diffusion
ViMo: A Generative Visual GUI World Model for App Agents
Dezhao Luo, Bohan Tang, Kang Li, Georgios Papoudakis, Jifei Song, Shaogang Gong, Jianye HAO, Jun Wang, Kun Shao
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add ViMo visual world model to T3A | T3A · step accuracy | step accuracy: 43.13 -> 49.20 | Improved |
| add ViMo visual world model to M3A | M3A · step accuracy | step accuracy: 46.01 -> 50.16 | Improved |
| add ViMo vs language-based world model Change-text | M3A · step accuracy | step accuracy: 47.28 -> 50.16 | Improved |
| add language-based HTML-text world model | M3A · step accuracy | step accuracy: 46.01 -> 46.65 | No clear effect |
| add ViMo in online navigation | T3A (Gemini-2.0-Flash) · AndroidWorld task accuracy | task accuracy: 33.19 -> 40.95 | Improved |
| add ViMo to unseen Apps zero-shot | M3A (Gemini-2.0-Flash) · step accuracy | step accuracy: 44.2 -> 47.6 | Improved |
| decouple graphic (STR) from text generation | ViMo vs IP2P* · GUI quality automatic | s_h: 0.69 -> 0.76 | Improved |
| ViMo vs HTML-vision on instructional accuracy | HTML-vision · GUI quality automatic | s_ia: 85.77 -> 75.39 | Hurt |
| ViMo vs IP2P* in user study | IP2P* · GUI quality user study | s_h: 0.63 -> 0.88 | Improved |
| retain static text as pixels; use action instruction conditioning (ablation) | ViMo · step accuracy | step accuracy | Improved |
| increase rollout iterations beyond one (ablation) | ViMo · step accuracy | step accuracy | Mixed |
| longer trajectories (T+1 to T+4) | ViMo · trajectory synthesis | trajectory accuracy: 57.89 -> 12.28 | Hurt |
Closest papers by meaning
- Navigating the Digital World as Humans Do: Universal Visual Grounding for GUI Agents (ICLR 2025 · similarity 0.89)
- ShowUI: One Vision-Language-Action Model for GUI Visual Agent (CVPR 2025 · similarity 0.89)
- Generative Visual Code Mobile World Models (ICML 2026 · similarity 0.89)
- UI-Vision: A Desktop-centric GUI Benchmark for Visual Perception and Interaction (ICML 2025 · similarity 0.88)
- Code2World: A GUI World Model via Renderable Code Generation (NeurIPS 2026 · similarity 0.88)
- OS-ATLAS: Foundation Action Model for Generalist GUI Agents (ICLR 2025 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
