ResearchDatasetsDynSess-EvalDynSess-EvalFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsRole-playingDynSess-Eval Human Evaluation 1.0 (test)3.38Average Score10Evaluator Alignment with Human JudgmentsDynSess-Eval 8 personas (test)67Role Consistency Rank Acc7