ResearchBenchmarksCounterfactual Fairness on DialogSum (DS)Follow0.1Sentiment ParityGem-Pro0.0640.3070.550.793Jul 15, 2024Evaluation ResultsMethodMethodLinksSentiment ParityC-ROUGE-LC-BLEUC-Cosine SimilarityGem-ProModel=Gem-ProModel=Gem-Pro2024.070.159.838.489.7GPT-4o-mModel=GPT-4o-mModel=GPT-4o-m2024.070.264.446.691.1Gem-FlModel=Gem-FlModel=Gem-Fl2024.070.558.536.289.1GPT-4oModel=GPT-4oModel=GPT-4o2024.070.959.439.390.4Gem-Fl-LtModel=Gem-Fl-LtModel=Gem-Fl-Lt2024.07161.441.990