ResearchTasksKnowledge-based Agent ReasoningFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedKARLBench 1.0 (test)Claude 4.6 Opus75.9BrowseComp-Plus Score17Mar 6, 2026