ResearchDatasetsEndoAgentBenchEndoAgentBenchFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsHigh-level Diagnostic ReasoningEndoAgentBench76.03CAP (CAR)8Fine-grained PerceptionEndoAgentBench75.29Localization Classification Accuracy8