ResearchTasksExpert-level Science Question AnsweringExpert-level Science Question AnsweringFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedGPQAUniSD*36.4Accuracy14May 8, 2026GPQA Diamondgpt-oss-120b78Score10Apr 23, 2026GPQA diamondMUR47.47Accuracy3May 12, 2026