Optimistic Meta-Gradients
Abstract
We study the connection between gradient-based meta-learning and convex optimisation. We observe that gradient descent with momentum is a special case of meta-gradients, and building on recent results in optimisation, we prove convergence rates for meta learning in the single task setting. While a meta-learned update rule can yield faster convergence up to constant factor, it is not sufficient for acceleration. Instead, some form of optimism is required. We show that optimism in meta-learning can be captured through the recently proposed Bootstrapped Meta-Gradient (Flennerhag et. al., 2022) method, providing deeper insight into its underlying mechanics.
Cite
Text
Flennerhag et al. "Optimistic Meta-Gradients." Neural Information Processing Systems, 2023.Markdown
[Flennerhag et al. "Optimistic Meta-Gradients." Neural Information Processing Systems, 2023.](https://mlanthology.org/neurips/2023/flennerhag2023neurips-optimistic/)BibTeX
@inproceedings{flennerhag2023neurips-optimistic,
title = {{Optimistic Meta-Gradients}},
author = {Flennerhag, Sebastian and Zahavy, Tom and O'Donoghue, Brendan and van Hasselt, Hado P and György, András and Singh, Satinder P.},
booktitle = {Neural Information Processing Systems},
year = {2023},
url = {https://mlanthology.org/neurips/2023/flennerhag2023neurips-optimistic/}
}