ResearchTasksHigh-Level ReasoningHigh-Level ReasoningFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedMMT-47 High Level ReasoningMoEDoRA45.92Accuracy17Apr 6, 2026HLEOpenAI DeepResearch26.6Average Score17Feb 26, 2026HLEReAct29.7HLE Score (Text)8Jul 1, 2026