List your company

Mobile UI Control on Robustness Benchmark

43.3LROS-Atlas-7B
9.70818.42927.1535.871Apr 7, 2026

Evaluation results

MethodLinks
OS-Atlas-7BCategory=Open-source 7B/8B2026.0443.310.7
Qwen2.5-VL-3BCategory=Open-source 3B2026.043035
UI-R1-3BCategory=Open-source 3B2026.0429.529
VeriGUI-3BCategory=Ours2026.0424.351.1
Qwen2.5-VL-7BCategory=Open-source 7B/8B2026.0423.534.1
UI-S1-7BCategory=Open-source 7B/8B2026.0420.543.6
VeriGUI-7BCategory=Ours2026.0415.652.5
Gemini-3-flashCategory=Closed-source2026.041537
UI-TARS-7BCategory=Open-source 7B/8B2026.0413.445.5
GPT-5.1Category=Closed-source2026.041121.9