ResearchBenchmarksSummary Similarity Evaluation on Flash generated summaries 2.5Follow0.872BERTScore F1AutoMUP0.848080.854290.86050.86671Apr 8, 2026Evaluation ResultsMethodMethodLinksBERTScore F1SBERT SimilaritySimCSE SimilarityUSE SimilarityROUGE-LBLEURT ScoreAutoMUPConsensus level=A1Consensus level=A12026.040.8720.720.9750.71124.640.5AutoMUPConsensus level=A2Consensus level=A22026.040.860.6340.9730.6616.630.9AutoMUPConsensus level=A3Consensus level=A32026.040.8490.6140.9690.6314.425.7