ResearchDatasets10-task benchmark10-task benchmarkFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsReward rollout alignment10-task benchmark T1: Folding Shorts0.889Rollout ρ8Reward prediction10-task benchmark S1 classic0.006Demo L (MSE)8Reward rollout alignment10-task benchmark T2: Cleaning Whiteboard0.667Rollout ρ7