ResearchBenchmarksExploration on ALFWorld (val)Follow97.9Success RateMemRL83.02886.88990.7594.611Jan 6, 2026Evaluation ResultsMethodMethodLinksSuccess RateMemRLModel=GPT-5-miniModel=GPT-5-mini2026.0197.9RAGModel=GPT-5-miniModel=GPT-5-mini2026.0195Self-RAGModel=GPT-5-miniModel=GPT-5-mini2026.0195Mem0Model=GPT-5-miniModel=GPT-5-mini2026.0195MemPModel=GPT-5-miniModel=GPT-5-mini2026.0192.1No MemoryModel=GPT-5-miniModel=GPT-5-mini2026.0183.6