ResearchDatasetsDiffReactFollowBenchmarksTask NameDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyTask NameDataset NameSOTA ResultTrendResultsContext window selectionDiffReact PDEBench (full-oracle case-level)5Lsel24Regret assessmentDiffReact full-oracle case-level PDEBench-0.8Regret Knee16Trajectory Rollout PredictionDiffReact f2 dense (OOD)0.086OOD Mean Error5