ResearchTasksSycophancy ResistanceSycophancy ResistanceFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedAnthropic model-written-evals sycophancy suiteInstruction Tuned94.6Overall Sycophancy Resistance Score6Jun 5, 2026Held-out MCQ sycophancy benchmark (aggregate)Instruction Tuned74MCQ Aggregate Score6Jun 5, 2026
Anthropic model-written-evals sycophancy suiteInstruction Tuned94.6Overall Sycophancy Resistance Score6Jun 5, 2026