ResearchBenchmarksHigh-level planning on Simulated Tasks Medium 3–7 actionsFollow97.96Success RateGemini91.4693.147594.83596.5225Oct 13, 2024Evaluation ResultsMethodMethodLinksSuccess RateGeminiPlanning LLM=Gemini, D...Planning LLM=Gemini, DINO Skill Discovery Model=Gemini2024.1097.96GPTPlanning LLM=GPT, DINO...Planning LLM=GPT, DINO Skill Discovery Model=Gemini2024.1097.71GPTPlanning LLM=GPT, DINO...Planning LLM=GPT, DINO Skill Discovery Model=None2024.1092.29GPTPlanning LLM=GPT, DINO...Planning LLM=GPT, DINO Skill Discovery Model=GPT2024.1091.71