“Meta-Learners’ Learning Dynamics Are unlike Learners’”, 2019-05-03 (; similar):
Meta-learning is a tool that allows us to build sample-efficient learning systems.
Here we show that, once meta-trained, LSTM meta-learners aren’t just faster learners than their sample-inefficient deep learning (DL) and reinforcement learning (RL) brethren, but that they actually pursue fundamentally different learning trajectories.
We study their learning dynamics on 3 sets of structured tasks for which the corresponding learning dynamics of DL and RL systems have been previously described: linear regression ( et al 2013), nonlinear regression ( et al 2018; et al 2018), and contextual bandits ( et al 2019).
In each case, while sample-inefficient DL and RL Learners uncover the task structure in a staggered manner, meta-trained LSTM meta-learners uncover almost all task structure concurrently, congruent with the patterns expected from Bayes-optimal inference algorithms.
This has implications for research areas wherever the learning behavior itself is of interest, such as safety, curriculum design, and human-in-the-loop machine learning.