ResearchTasksAutonomous replicationAutonomous replicationFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedtarget-1 (citrusdrop) API model runsGPT-5.185Non-refusal Count6May 11, 2026target-1 citrusdropQwen3.6-27B48Non-refusal Rate4May 11, 2026