ResearchTasksSelf-confidence estimationSelf-confidence estimationFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedAIME hard OOD 23–25 (test)OCB12.9Accuracy6Feb 26, 2026GSM8K easy OOD (test)RLCR89.8Accuracy6Feb 26, 2026MATH500 in-distribution (test)RLCR78.6Accuracy6Feb 26, 2026