TY  - RPRT
TI  - Reinforcement Learning as (Discrete) Potential Theory
AU  - Christopher Connolly
PY  - 2026
UR  - https://arxiv.org/abs/2608.17181
ID  - 2608.17181
ER  -