ResearchTasksEmbodied Agent TaskEmbodied Agent TaskFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedALFWorld UnseenGRPO w/ EVU79.1Success Rate40Jun 1, 2026ScienceWorld UnseenGRPO w/ EVU70.8Success Rate18Apr 21, 2026ScienceWorld SeenGRPO w/ EVU70.9Success Rate18Apr 21, 2026VirtualHome UnseenGRPO w/ EVU36.9Success Rate18Apr 21, 2026ALFWorld SeenGRPO w/ EVU85.7Success Rate (%)18Apr 21, 2026