BIMRL: Brain Inspired Meta Reinforcement Learning

TL;DR

BIMRL introduces a brain-inspired meta-reinforcement learning architecture with a novel memory module and intrinsic reward, enabling agents to adapt quickly to new tasks and outperform existing methods in MiniGrid environments.

Contribution

The paper presents a new multi-layer architecture with a brain-inspired memory module and intrinsic reward for improved meta-RL performance.

Findings

Outperforms strong baselines on MiniGrid environments

Effective in rapid adaptation to new tasks

Memory module enhances exploration and learning efficiency

Abstract

Sample efficiency has been a key issue in reinforcement learning (RL). An efficient agent must be able to leverage its prior experiences to quickly adapt to similar, but new tasks and situations. Meta-RL is one attempt at formalizing and addressing this issue. Inspired by recent progress in meta-RL, we introduce BIMRL, a novel multi-layer architecture along with a novel brain-inspired memory module that will help agents quickly adapt to new tasks within a few episodes. We also utilize this memory module to design a novel intrinsic reward that will guide the agent's exploration. Our architecture is inspired by findings in cognitive neuroscience and is compatible with the knowledge on connectivity and functionality of different regions in the brain. We empirically validate the effectiveness of our proposed method by competing with or surpassing the performance of some strong baselines on multiple MiniGrid environments.