Xiaokang(Koe) Ye(@koe_ye40329):A lot of the work behind DeliveryGym is systems side: making long-horizon agentic online RL practical in UE5 while balancing photorealistic rendering with rollout throughput. We use SPEAR for the Python interface to Unreal, alongside our distributed rollout infra: parallel simulator workers, asynchronous coordination across environment execution, inference, and policy training. Thanks @mikeroberts3000 and the SPEAR team for building and sharing these tools!

Xiaokang(Koe) Ye

@koe_ye40329

Ph.D. student @UCSanDiego. Leading @simworld_ai.

Joined February 2024

269 Following 168 Followers

Xiaokang(Koe) Ye@koe_ye40329

2026.09.21 12:32

A lot of the work behind DeliveryGym is systems side: making long-horizon agentic online RL practical in UE5 while balancing photorealistic rendering with rollout throughput. We use SPEAR for the Python interface to Unreal, alongside our distributed rollout infra: parallel simulator workers, asynchronous coordination across environment execution, inference, and policy training. Thanks @mikeroberts3000 and the SPEAR team for building and sharing these tools!