Learning in Low-Dimensional Subspaces: Orthogonal Bottlenecks for Reinforcement Learning

ArXi:2605.26012v1 Announce Type: cross Deep reinforcement learning (RL) agents commonly rely on high-dimensional neural representations, despite growing evidence that task-relevant value and policy structure may be intrinsically low-dimensional. In this work, we present a simple yet effective representation-level prior that inserts a fixed orthonormal projection to constrain encoder features to a low-dimensional subspace, requiring no auxiliary objectives, pre