Each LLM is in an independent loop, where each turn they are given the current speed and time to impact, as well as their historical latency. They can pre-plan moves to account for latency.
Each LLM is in an independent loop, where each turn they are given the current speed and time to impact, as well as their historical latency. They can pre-plan moves to account for latency.