MoSE: Skill-by-Skill Mixture-of-Experts Learning for Embodied Autonomous Machines

TL;DR

MoSE introduces a skill-by-skill mixture-of-experts approach for embodied AI, improving reasoning and learning efficiency in autonomous systems by mimicking human skill learning and reasoning processes.

Contribution

The paper proposes a novel skill-oriented MoE model with a hierarchical dataset and routing mechanism, enabling efficient skill-by-skill learning in embodied AI tasks.

Findings

Outperforms existing models on autonomous driving reasoning tasks

Achieves better results with less than 40% of parameters

Effectively integrates auxiliary tasks without extra computational cost

Abstract

To meet the growing demand for smarter, faster, and more efficient embodied AI solutions, we introduce a novel Mixture-of-Expert (MoE) method that significantly boosts reasoning and learning efficiency for embodied autonomous systems. General MoE models demand extensive training data and complex optimization, which limits their applicability in embodied AI such as autonomous driving (AD) and robotic manipulation. In this work, we propose a skill-oriented MoE called MoSE, which mimics the human learning and reasoning process skill-by-skill, step-by-step. We introduce a skill-oriented routing mechanism that begins with defining and annotating specific skills, enabling experts to identify the necessary competencies for various scenarios and reasoning tasks, thereby facilitating skill-by-skill learning. To better align with multi-step planning in human reasoning and in end-to-end driving models, we build a hierarchical skill dataset and pretrain the router to encourage the model to think step-by-step. Unlike other multi-round dialogues, MoSE integrates valuable auxiliary tasks (e.g. perception-prediction-planning for AD, and high-level and low-level planning for robots) in one single forward process without introducing any extra computational cost. With less than 3B sparsely activated parameters, our model effectively grows more diverse expertise and outperforms models on both AD corner-case reasoning tasks and robot reasoning tasks with less than 40% of the parameters.