ResearchTasksRuntime monitor executionFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedLabUtopia L1-L4 grounding 1.0 (source-held-out)LabGuard-Hybrid23.8Unsafe Events10Jul 1, 2026LabUtopia L1-L4 chemistry-held-out grounding 1.0 (hydrochloric acid)LabGuard-Hybrid14.3Unsafe Events6Jul 1, 2026
LabUtopia L1-L4 chemistry-held-out grounding 1.0 (hydrochloric acid)LabGuard-Hybrid14.3Unsafe Events6Jul 1, 2026