ResearchTasksReasoning Efficiency EvaluationFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedQwen3 Reasoning Tasks Five-component aggregate BaseRollout-level |Ai| replay0.71Token-based AES15Jun 4, 2026
Qwen3 Reasoning Tasks Five-component aggregate BaseRollout-level |Ai| replay0.71Token-based AES15Jun 4, 2026