TY  - RPRT
TI  - StructRL: Recovering Dynamic Programming Structure from Learning Dynamics in Distributional Reinforcement Learning
AU  - Ivo Nowak
PY  - 2026
UR  - https://arxiv.org/abs/2604.08620
ID  - 2604.08620
ER  -