TY  - RPRT
TI  - Learning in Low-Dimensional Subspaces: Orthogonal Bottlenecks for Reinforcement Learning
AU  - Aleksandar Todorov
AU  - Matthia Sabatelli
PY  - 2026
UR  - https://arxiv.org/abs/2605.26012
ID  - 2605.26012
ER  -