ResearchBenchmarksParaphrase Semantic Similarity on Human Evaluation 300 samples (random sample)For AI assistantsMarkdownContent indexParaphrase Semantic Similarity on Human Evaluation 300 samples (random sample)Follow28.7Score 3PARANMT26.41227.00627.628.194May 26, 2023Evaluation resultsMethodMethodLinksScore 3Score 2Score 1Average ScorePARANMT2023.0528.746.724.62.04PARABANK12023.0526.84924.22.03PARABANK22023.0526.850.322.92.04PARAAMR2023.0526.547.226.32