Global Optimality in Bivariate Gradient-based DAG Learning | Read Paper on Bytez