_NeurIPS_2022__On_the_Effectiveness_of_Fine_tuning_Versus_Meta_reinforcement_Learning (1)

Meta-learning approaches have emerged as a popular solution to achieve this. However, meta-reinforcement learning (meta-RL) algorithms have thus far been predominately validated on simple environments with narrow task distributions.