BADGER: Learning to (Learn [Learning Algorithms] through Multi-Agent Communication)

Abstract

In this work, we propose a novel memory-based multi-agent meta-learning architecture and learning procedure that allows for learning of a shared communication policy that enables the emergence of rapid adaptation to new and unseen environments by learning to learn learning algorithms through communication. Behavior, adaptation and learning to adapt emerges from the interactions of homogeneous experts inside a single agent. The proposed architecture should allow for generalization beyond the level seen in existing methods, in part due to the use of a single policy shared by all experts within the agent as well as the inherent modularity of 'Badger'.

References in corpus (12)

• Learning to reinforcement learn
• One Model To Learn Them All
• Paired Open-Ended Trailblazer (POET): Endlessly Generating Increasingly Complex and Diverse Learning Environments and Their Solutions
• Learning to Learn: Meta-Critic Networks for Sample Efficient Learning
• Causal Reasoning from Meta-reinforcement Learning
• Autocurricula and the Emergence of Innovation from Social Interaction: A Manifesto for Multi-Agent Intelligence Research
• Improving Generalization in Meta Reinforcement Learning using Learned Objectives
• Meta-learning of Sequential Strategies
• Meta-learners' learning dynamics are unlike learners'
• Learning to Communicate in Multi-Agent Reinforcement Learning : A Review
• Reinforcement Learning applied to Single Neuron
• Modular meta-learning in abstract graph networks for combinatorial generalization