ResearchDatasetsOpenClawBenchFollowBenchmarksTask NameDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyTask NameDataset NameSOTA ResultTrendResultsVulnerability DetectionOpenClawBench held-out set n=217 (test)75.1Recall4Vulnerability DetectionOpenClawBench rule-derivation set n=229 (train)79Recall4