Disentangled Skill Embeddings for Reinforcement Learning
This addresses the problem of efficient skill learning and transfer in reinforcement learning for agents, though it appears incremental as it builds on existing variational inference and multi-task methods.
The paper tackles multi-task reinforcement learning by proposing Disentangled Skill Embeddings, a framework that learns policies generalizing across changing dynamics and goals, enabling transfer and specialization with task-specific latent-space embeddings, and shows generalization to unseen conditions while serving as skills for hierarchical reinforcement learning.
We propose a novel framework for multi-task reinforcement learning (MTRL). Using a variational inference formulation, we learn policies that generalize across both changing dynamics and goals. The resulting policies are parametrized by shared parameters that allow for transfer between different dynamics and goal conditions, and by task-specific latent-space embeddings that allow for specialization to particular tasks. We show how the latent-spaces enable generalization to unseen dynamics and goals conditions. Additionally, policies equipped with such embeddings serve as a space of skills (or options) for hierarchical reinforcement learning. Since we can change task dynamics and goals independently, we name our framework Disentangled Skill Embeddings (DSE).