POPGym Arcade: Parallel Pixelated POMDPs

원문 초록

We present the POPGym Arcade, a collection of hardware-accelerated, pixel-based environments with shared observation and action spaces. Each environment includes fully and partially observable variants, enabling counterfactual studies on partial observability. We also introduce mathematical tools for analyzing policies under partial observability, which reveal how agents recall past information to make decisions. Our analysis shows (1) that controlling for partial observability is critical and (2) that agents with long-term memory learn brittle policies that struggle to generalize. Finally, we demonstrate that recurrent policies can be "poisoned" by old, out-of-distribution observations, with implications for sim-to-real transfer, imitation learning, and offline reinforcement learning.