A Policy Gradient Algorithm for Learning to Learn
              in Multiagent Reinforcement Learning

     Dong-Ki Kim 1 2 Miao Liu 2 3 Matthew Riemer 2 3 Chuangchuang Sun 1 2 Marwa Abdulhai 1 2
        Golnaz Habibi 1 2 Sebastian Lopez-Cot 1 2 Gerald Tesauro 2 3 Jonathan P. How 1 2

            Abstract                 learning agents because their changing behaviors jointly af-
   A fundamental challenge in multiagent reinforce-      fect the environment’s transition and reward function. This
...