ResearchTasksSocial Interaction EvaluationSocial Interaction EvaluationFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedSOTOPIA-Hard GPT-4o-as-PartnerAMPO7.68Goal Score24Mar 4, 2026SOTOPIA GPT-4o-as-PartnerAMPO8.75Goal Score24Mar 4, 2026SOTOPIA-Hard (Self-Play)AMPO8.06GOAL Score24Mar 4, 2026SOTOPIA (Self-Play)AMPO9.08Goal Score24Mar 4, 2026SOTOPIAAURA9Believability Score1Jun 5, 2026