ResearchTasksOptical ReasoningFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedAquaRat, Gsm8k, GPQA Diamond, ScienceQA, Zebra-Cot (test)Claude 4.5 (Text reasoning)84.65Accuracy (AquaRat)35Jun 9, 2026
AquaRat, Gsm8k, GPQA Diamond, ScienceQA, Zebra-Cot (test)Claude 4.5 (Text reasoning)84.65Accuracy (AquaRat)35Jun 9, 2026