ResearchTasksAbstention and Hallucination EvaluationFor AI assistantsMarkdownContent indexAbstention and Hallucination EvaluationFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedControlled 50-item Evaluation set epistemic regimes R1–R5Instruction98Accuracy12Apr 9, 2026