ResearchTasksPrivacy-Aware Minimal View for AgentsFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedAugmented WebArena N = 662 (test)Llama-3.3-70B-Instruct96.6Context Recall (TCNP)8Jun 15, 2026