On the grid-sampling limit SDE

Christian Bender, Nguyen Tran Thuan

In our recent work [3] we introduced the grid-sampling SDE as a proxy for modeling exploration in continuous-time reinforcement learning. In this note, we provide further motivatio…