ResearchDatasetsLLaMA-2 generated sequencesLLaMA-2 generated sequencesFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsWatermark Robustness under GPT rephrasing attackLLaMA-2 generated sequences (1,000)62.5TPR@FPR=5%7