ResearchDatasetsMercuryFollowBenchmarksTask NameDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyTask NameDataset NameSOTA ResultTrendResultsCode EfficiencyMercury83.6Beyond@129Out-of-distribution AI-generated text detectionMercury Out-of-distribution (OOD) 2 (test unseen domains)95.7Legal Accuracy16Code GenerationMercury (test)90.3Pass Rate (Easy)8Black-box Vulnerability DetectionMercury22Vulnerabilities Found Count4
Out-of-distribution AI-generated text detectionMercury Out-of-distribution (OOD) 2 (test unseen domains)95.7Legal Accuracy16