ResearchTasksContext LearningFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedDeR2 BenchGPT-5.2 High69Score12Jul 2, 2026ToB Complex WorkflowsGemini-3-pro High69.2Score8Jul 2, 2026ToB-Reference Q&ASeed2.0 Pro72.4Score8Jul 2, 2026CL-BenchGPT-5-mini High25.2Score8Jul 2, 2026KORBenchGPT-5.2 High79.2Score8Jul 2, 2026CL-Bench (test)Context-CoT SFT12.85Overall Score8May 26, 2026