ResearchBenchmarksSpatial Control (right_of) on Spatial Control TasksFollow85Success RateLLM+DL5965.7572.579.25Dec 19, 2025Evaluation ResultsMethodMethodLinksSuccess RateAverage StepsLLM+DLLLM Model=PhiLLM Model=Phi2025.12850.85LLM+DLLLM Model=LLaMA-3.2LLM Model=LLaMA-3.22025.12751.35DL-onlyLLM Model=NoneLLM Model=None2025.12701.35LLM-onlyLLM Model=MistralLLM Model=Mistral2025.12703.6LLM+DLLLM Model=MistralLLM Model=Mistral2025.12701.15LLM-onlyLLM Model=LLaMA-3.2LLM Model=LLaMA-3.22025.12703.4LLM-onlyLLM Model=PhiLLM Model=Phi2025.12604.95