Deterministic Policy Optimization by Combining Pathwise and Score Function Estimators for Discrete Action Spaces | Read Paper on Bytez