ResearchBenchmarksRobot Task Planning on RearrangementFollow212Token Cost (k)LLMTAMP162.08499.048361,172.96May 23, 2025Evaluation ResultsMethodMethodLinksToken Cost (k)Success RateLLMTAMPLLM Backbone=GPT-4oLLM Backbone=GPT-4o2025.05212—PDDLLMExecution=PDDL solverExecution=PDDL solver2025.05334—O1-TAMPLLM Backbone=OpenAI o1LLM Backbone=OpenAI o12025.051,200—R1-TAMPLLM Backbone=DeepSeek R1LLM Backbone=DeepSeek R12025.051,460—