Adaptive Reinforcement Learning through Evolving Self-Modifying Neural Networks

TL;DR

This paper introduces self-modifying neural networks that adapt online in reinforcement learning, enabling better and faster adaptation in complex tasks like quadruped locomotion with limb failures.

Contribution

It presents a novel approach of evolving self-modifying neural networks for online adaptation, outperforming traditional gradient-based methods in complex meta-learning scenarios.

Findings

Self-modifying networks adapt more effectively to limb failures.

Evolved networks outperform gradient-based updates in complex tasks.

Training time is reduced with self-modifying plastic networks.

Abstract

The adaptive learning capabilities seen in biological neural networks are largely a product of the self-modifying behavior emerging from online plastic changes in synaptic connectivity. Current methods in Reinforcement Learning (RL) only adjust to new interactions after reflection over a specified time interval, preventing the emergence of online adaptivity. Recent work addressing this by endowing artificial neural networks with neuromodulated plasticity have been shown to improve performance on simple RL tasks trained using backpropagation, but have yet to scale up to larger problems. Here we study the problem of meta-learning in a challenging quadruped domain, where each leg of the quadruped has a chance of becoming unusable, requiring the agent to adapt by continuing locomotion with the remaining limbs. Results demonstrate that agents evolved using self-modifying plastic networks are more capable of adapting to complex meta-learning learning tasks, even outperforming the same network updated using gradient-based algorithms while taking less time to train.