ResearchBenchmarksMulti-turn consistency evaluation on MT-Consistency (test)Follow71.34Acc AvgGPT40.327248.378656.4364.4814Mar 28, 2025Evaluation ResultsMethodMethodLinksAcc AvgR_swayPWCScoreGPTSetting=Multi-TurnSetting=Multi-Turn2025.0371.346.841.69ClaudeSetting=Multi-TurnSetting=Multi-Turn2025.0363.074.381.51QwenSetting=Multi-TurnSetting=Multi-Turn2025.0360.866.021.64MistralSetting=Multi-TurnSetting=Multi-Turn2025.0350.025.281.53GeminiSetting=Multi-TurnSetting=Multi-Turn2025.0341.843.881.25LlaMaSetting=Multi-TurnSetting=Multi-Turn2025.0341.524.591.45