ResearchBenchmarksToxicity Evaluation on DialogSum (DS)Follow0Toxic FractionGPT-4o-0.0080.0460.10.154Jul 15, 2024Evaluation ResultsMethodMethodLinksToxic FractionGPT-4o2024.070GPT-4o-m2024.070Gem-Fl2024.070.1Gem-Pro2024.070.1Gem-Fl-Lt2024.070.2