ResearchDatasetsAnthropic HarmlessAnthropic HarmlessFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsReward ModelingAnthropic Harmless74.7Accuracy12Reward Model TransferAnthropic Harmless (AHar)4.62AOG8Open-ended DialogueAnthropic-Harmless ID68.1Win Rate4