ResearchDatasetsBALROGBALROGFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsLong-horizon game playingBALROG NetHack6.8Progression Rate12Long-horizon game playingBALROG MiniHack40Progression Rate12Agentic reasoningBALROG31.5Accuracy8