ResearchDatasetsARBARBFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsReasoningARB OOD88.7Accuracy10ReasoningARB ID91.2Accuracy (ARB ID)10Information RetrievalARB internal deployment (val)0Observed False Positives6Information RetrievalARB Sem. Relevant (val)99.67F1 Score5