ResearchBenchmarksRobot Task Planning on Bridge BuildingFollow50Token Cost (k)LLMTAMP37124.75212.5300.25May 23, 2025Evaluation ResultsMethodMethodLinksToken Cost (k)Success RateLLMTAMPLLM Backbone=GPT-4oLLM Backbone=GPT-4o2025.0550—O1-TAMPLLM Backbone=OpenAI o1LLM Backbone=OpenAI o12025.05270—R1-TAMPLLM Backbone=DeepSeek R1LLM Backbone=DeepSeek R12025.05363—PDDLLMExecution=PDDL solverExecution=PDDL solver2025.05375—