ResearchBenchmarksDialogue Simulation on DialSimFollow4.12F1 ScoreLightMem1.06241.85622.653.4438Apr 9, 2026Evaluation ResultsMethodMethodLinksF1 ScoreBLEU-1 ScoreROUGE-L ScoreROUGE-2 ScoreMETEOR ScoreSBERT ScoreLightMemBackbone=GPT-4o-miniBackbone=GPT-4o-mini2026.044.123.954.24.152.4823.4A-MEMBackbone=GPT-4o-miniBackbone=GPT-4o-mini2026.043.453.373.543.62.0519.51LoCoMoBackbone=GPT-4o-miniBackbone=GPT-4o-mini2026.042.553.132.750.91.6415.76MemGPTBackbone=GPT-4o-miniBackbone=GPT-4o-mini2026.041.181.070.960.420.958.54