ResearchTasksMultiple Choice QuestionFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedPhyX mini (test)GRPO (Fmt + Acc + ASM)46.2Overall Accuracy7Apr 16, 2026NegBench COCO subsetG-DINO-B + NEGTOME32.55Overall Accuracy4Mar 24, 2026CICERO v2DIALECT-Large88.63Macro F12Feb 26, 2026CICERODIALECT-Large70.66Macro F12Feb 26, 2026