Recurrent Memory Decision Transformer

Transformative models, originally developed for natural language problems, have recently been widely used in offline reinforcement learning tasks. This is due to the fact that the agent's history can be represented as a sequence, and the whole task can be reduced to the sequence modeling task. However, the quadratic complexity of the transformer operation limits the potential increase in context. Therefore, to work with long sequences in a natural language, different versions of the memory mechanism are used. In this paper, we propose the Recurrent Memory Decision Transformer (RMDT), a model that uses a recurrent memory mechanism for reinforcement learning problems. We conduct thorough experiments on Atari games and MoJoCo control problems, and show that our proposed model is significantly superior to its counterparts without the recurrent memory mechanism on Atari games. We also carefully study the effect of memory on the performance of the proposed model. These findings shed light on the potential of incorporating recurrent memory mechanisms to improve the performance of large-scale transformer models in offline reinforcement learning tasks. The Recurrent Memory Decision Transformer code is publicly available in repository \url{https://anonymous.4open.science/r/RMDT-4FE4}.