Linear memory networks excel at partially observable reinforcement learning

Original: Why Linear Recurrent Memory Works in Partially Observable Reinforcement Learning

Source: arxiv.org ↗