TY - RPRT TI - A Differential and Pointwise Control Approach to Reinforcement Learning AU - Minh Nguyen AU - Chandrajit Bajaj PY - 2026 UR - https://arxiv.org/abs/2404.15617 ID - 2404.15617 ER -