Bootstrapped Meta-Learning

Meta-learning empowers artificial intelligence to increase its efficiency by learning how to learn. Unlocking this potential involves overcoming a challenging meta-optimisation problem. We propose an algorithm that tackles this problem by letting the <PRE_TAG>meta-learner</POST_TAG> teach itself. The algorithm first bootstraps a target from the <PRE_TAG>meta-learner</POST_TAG>, then optimises the <PRE_TAG>meta-learner</POST_TAG> by minimising the distance to that target under a chosen (pseudo-)metric. Focusing on meta-learning with gradients, we establish conditions that guarantee performance improvements and show that the metric can control meta-optimisation. Meanwhile, the <PRE_TAG>bootstrapping</POST_TAG> mechanism can extend the effective meta-learning horizon without requiring backpropagation through all updates. We achieve a new state-of-the art for model-free agents on the Atari ALE benchmark and demonstrate that it yields both performance and efficiency gains in multi-task meta-learning. Finally, we explore how <PRE_TAG>bootstrapping</POST_TAG> opens up new possibilities and find that it can meta-learn efficient exploration in an epsilon-greedy Q-learning agent, without backpropagating through the update rule.