ResearchTasksInteractive LLM Agent Task SuccessFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedALFWorld UnseenQ-Evolve90.3Success Rate7Jun 8, 2026ALFWorld SeenQ-Evolve89.6Success Rate7Jun 8, 2026SciWorld UnseenQ-Evolve82.4Success Rate7Jun 8, 2026SciWorld (Seen)Q-Evolve86.4Success Rate7Jun 8, 2026