Decision Mamba Architectures

TL;DR

This paper introduces Decision Mamba and Hierarchical Decision Mamba architectures, which enhance Transformer-based imitation learning models, outperforming existing methods across various environments and demonstrating significant improvements in task performance.

Contribution

The paper presents two novel Mamba architectures that improve upon Transformer models for imitation learning, with extensive experiments validating their effectiveness.

Findings

Mamba models outperform Transformers in most tasks

Decision Mamba outperforms other methods in diverse settings

Extensive experiments across OpenAI Gym and D4RL environments

Abstract

Recent advancements in imitation learning have been largely fueled by the integration of sequence models, which provide a structured flow of information to effectively mimic task behaviours. Currently, Decision Transformer (DT) and subsequently, the Hierarchical Decision Transformer (HDT), presented Transformer-based approaches to learn task policies. Recently, the Mamba architecture has shown to outperform Transformers across various task domains. In this work, we introduce two novel methods, Decision Mamba (DM) and Hierarchical Decision Mamba (HDM), aimed at enhancing the performance of the Transformer models. Through extensive experimentation across diverse environments such as OpenAI Gym and D4RL, leveraging varying demonstration data sets, we demonstrate the superiority of Mamba models over their Transformer counterparts in a majority of tasks. Results show that DM outperforms other methods in most settings. The code can be found at https://github.com/meowatthemoon/DecisionMamba.