ResearchDatasetsUROBenchUROBenchFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsTime-controlled instruction-followingUROBench long (30s–1min)6.13MAE8Time-controlled instruction-followingUROBench short (10s–30s)3.71MAE8