ResearchTasksUtility assessmentUtility assessmentFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedAgentDojo (test)Semantic-Injection100Utility128Feb 26, 2026MMLU-ProIdentity-Robust Generation16.5Personalization Bias (PB)9Feb 26, 2026LLaVA-Bench In-the-WildBase7.88Utility Score4Apr 20, 2026Financial Agent DialoguesFinSec90.13Utility F14Apr 13, 2026