ResearchTasksClick-in-targetFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedOOD Evaluation Suite (ScreenSpot-Pro, OSWorld-G, ScreenSpot-V2) unweighted meanQwen3.5-4B (base)60Click-in-Target Accuracy (SS-Pro)11Jun 25, 2026WinDOM in-distribution 2,000-example subsample (held-out)SFT82.1Click-in-target Accuracy11Jun 25, 2026
OOD Evaluation Suite (ScreenSpot-Pro, OSWorld-G, ScreenSpot-V2) unweighted meanQwen3.5-4B (base)60Click-in-Target Accuracy (SS-Pro)11Jun 25, 2026
WinDOM in-distribution 2,000-example subsample (held-out)SFT82.1Click-in-target Accuracy11Jun 25, 2026