ResearchDatasetsMultiNews, Qasper, RepoBench-P, and RULERMultiNews, Qasper, RepoBench-P, and RULERFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsLong-context evaluationMultiNews, Qasper, RepoBench-P, and RULER Averaged 128K (test)15.3Memory Footprint (GB)6
Long-context evaluationMultiNews, Qasper, RepoBench-P, and RULER Averaged 128K (test)15.3Memory Footprint (GB)6