ResearchTasksInteractive Agent Task CompletionFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedSocialMazeOpenSkill82.7Average Reward14Jun 8, 2026ScienceWorldOpenSkill90Average Reward12Jun 8, 2026ALFWorldrerank-SVMRank64.31Success Rate (SR)5Jun 19, 2026