Abstract:Standard reinforcement learning (RL) agents never intelligently explore like a human (i.e. taking into account complex domain priors and adapting quickly based on previous exploration). Across episodes, RL agents struggle to perform even simple exploration strategies, for example systematic search that avoids exploring the same location multiple times. This poor exploration limits performance on challenging domains. Meta-RL is a potential solution, as unlike standard RL, meta-RL can learn to explore, and potentially learn highly complex strategies far beyond those of standard RL, strategies such as experimenting in early episodes to learn new skills, or conducting experiments to learn about the current environment. Traditional meta-RL focuses on the problem of learning to optimally balance exploration and exploitation to maximize the cumulative reward of the episode sequence (e.g., aiming to maximize the total wins in a tournament -- while also improving as a player). We identify a new challenge with state-of-the-art cumulative-reward meta-RL methods. When optimal behavior requires exploration that sacrifices immediate reward to enable higher subsequent reward, existing state-of-the-art cumulative-reward meta-RL methods become stuck on the local optimum of failing to explore. Our method, First-Explore, overcomes this limitation by learning two policies: one to solely explore, and one to solely exploit. When exploring requires forgoing early-episode reward, First-Explore significantly outperforms existing cumulative meta-RL methods. By identifying and solving the previously unrecognized problem of forgoing reward in early episodes, First-Explore represents a significant step towards developing meta-RL algorithms capable of human-like exploration on a broader range of domains.

Unsupervised Meta-Learning for Reinforcement Learning

A Survey of Meta-Reinforcement Learning

Reinforcement Teaching

Guided Meta-Policy Search

Exploration With Task Information for Meta Reinforcement Learning

Meta-Learning in Neural Networks: A Survey

Model-based Adversarial Meta-Reinforcement Learning

Curriculum in Gradient-Based Meta-Reinforcement Learning

Meta-World: A Benchmark and Evaluation for Multi-Task and Meta Reinforcement Learning

Deep Online Learning via Meta-Learning: Continual Adaptation for Model-Based RL

First-Explore, then Exploit: Meta-Learning to Solve Hard Exploration-Exploitation Trade-Offs

Probabilistic Active Meta-Learning

Unsupervised Meta-Learning via In-Context Learning

MAML2: meta reinforcement learning via meta-learning for task categories

Revisiting Meta-Learning As Supervised Learning

A Meta Understanding of Meta-Learning

Rethinking Meta-Learning from a Learning Lens

Meta-Reinforcement Learning with Dynamic Adaptiveness Distillation

A Comprehensive Overview and Survey of Recent Advances in Meta-Learning

Making Scalable Meta Learning Practical

MAMBA: an Effective World Model Approach for Meta-Reinforcement Learning