ResearchTasksEgregious unfaithfulnessFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedAggregate of seven domains (including Moral Machines, Income, Breast Cancer Recurrence, and Pima Diabetes)Gemini 3 Flash5.8Egregious Unfaithfulness (%)10Feb 26, 2026
Aggregate of seven domains (including Moral Machines, Income, Breast Cancer Recurrence, and Pima Diabetes)Gemini 3 Flash5.8Egregious Unfaithfulness (%)10Feb 26, 2026