ResearchDatasetsAuto-ClawEvalFollowBenchmarksTask NameDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyTask NameDataset NameSOTA ResultTrendResultsAgent EvaluationAuto-ClawEval Mini (104 environments)94.2Safety Score8Agent EvaluationAuto-ClawEval93.3Safety8