ResearchBenchmarksTask Planning and Program Generation on IMR-Bench Complex Multi-RobotFollow88OCIMR-LLM48.4858.746979.26Mar 3, 2026Evaluation ResultsMethodMethodLinksOCSEExeGCRSRIMR-LLMBackbone=GPT-4oBackbone=GPT-4o2026.038875767968IMR-LLMBackbone=Qwen3-32BBackbone=Qwen3-32B2026.038571767960LiP-OBackbone=GPT-4oBackbone=GPT-4o2026.036328364224LaMMA-SBackbone=GPT-4oBackbone=GPT-4o2026.035626203316LaMMA-OBackbone=GPT-4oBackbone=GPT-4o2026.035626283720SMART-LLMBackbone=GPT-4oBackbone=GPT-4o2026.03504030