ResearchDatasetsVicunaEvalVicunaEvalFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsInstruction FollowingVicunaEval40.75VicunaEval Score80Instruction FollowingVicunaEval35Rouge-L72General PerformanceVicunaEval96.3Winrate21Instruction FollowingVicunaEval (OOD)19.7Rouge-L20GenerationVicunaEval (test)56.07LLM Judge Score2Instruction FollowingVicunaEval (test)—Score0