ResearchTasksAutomated agent rollbackFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedGSM8K Right Dense variant (test)LF75Success Rate3May 11, 2026GSM8K Mid Dense variant (test)LF68Success Rate3May 11, 2026GSM8K Left Dense variant (test)LF77Success Rate3May 11, 2026