ResearchTasksWatermark Robustness under GPT rephrasing attackFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedLLaMA-2 generated sequences (1,000)MCmark62.5TPR@FPR=5%7Feb 26, 2026