English

Hierarchical Transformers are Efficient Meta-Reinforcement Learners

Machine Learning 2024-02-12 v1 Artificial Intelligence

Abstract

We introduce Hierarchical Transformers for Meta-Reinforcement Learning (HTrMRL), a powerful online meta-reinforcement learning approach. HTrMRL aims to address the challenge of enabling reinforcement learning agents to perform effectively in previously unseen tasks. We demonstrate how past episodes serve as a rich source of information, which our model effectively distills and applies to new contexts. Our learned algorithm is capable of outperforming the previous state-of-the-art and provides more efficient meta-training while significantly improving generalization capabilities. Experimental results, obtained across various simulated tasks of the Meta-World Benchmark, indicate a significant improvement in learning efficiency and adaptability compared to the state-of-the-art on a variety of tasks. Our approach not only enhances the agent's ability to generalize from limited data but also paves the way for more robust and versatile AI systems.

Keywords

Cite

@article{arxiv.2402.06402,
  title  = {Hierarchical Transformers are Efficient Meta-Reinforcement Learners},
  author = {Gresa Shala and André Biedenkapp and Josif Grabocka},
  journal= {arXiv preprint arXiv:2402.06402},
  year   = {2024}
}
R2 v1 2026-06-28T14:44:02.859Z