Optimistic Meta-Gradients
Not on the map. Why?
Abstract (arXiv)
We study the connection between gradient-based meta-learning and convex op-timisation. 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 Bootstrapped Meta-Gradients (Flennerhag et al., 2022), providing deeper insight into its underlying mechanics.
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 27 of its 48 references that are in the map. First posted on arXiv 9 Jan 2023.
Builds on
27 of 48 references.
References · 27
- 2017Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networkscited by 4,861
- 2016Matching Networks for One Shot Learningcited by 2,246
- 2018On First-Order Meta-Learning Algorithmscited by 1,003
- 2020Meta-Learning in Neural Networks: A Surveycited by 944
- 2016Learning to learn by gradient descent by gradient descentcited by 616
- 2016Learning to reinforcement learncited by 369
- 2015Gradient-based Hyperparameter Optimization through Reversible Learningcited by 357
- 2013Optimization, Learning, and Games with Predictable Sequencescited by 187
- 2019Adaptive Gradient-Based Meta-Learning Methodscited by 183
- 2018Meta-Gradient Reinforcement Learningcited by 156
- 2019Online Meta-Learningcited by 137
- 2020AutoML-Zero: Evolving Machine Learning Algorithms From Scratchcited by 137
15 more
- 2019On the Convergence Theory of Gradient-Based Model-Agnostic Meta-Learning Algorithmscited by 132
- 2020Discovering Reinforcement Learning Algorithmscited by 75
- 2019Improving Generalization in Meta Reinforcement Learning using Learned Objectivescited by 68
- 2020Meta-Gradient Reinforcement Learning with an Objective Discovered Onlinecited by 52
- 2016MetaGrad: Multiple Learning Rates in Online Learningcited by 46
- 2021Bootstrapped Meta-Learningcited by 45
- 2020Meta-learning curiosity algorithmscited by 25
- 2017A Modular Analysis of Adaptive (Non-)Convex Optimization: Optimism, Composite Objectives, and Variational Boundscited by 23
- 2018Acceleration through Optimistic No-Regret Dynamicscited by 22
- 2018Transferring Knowledge across Learning Processescited by 22
- 2022Global Convergence of MAML and Theory-Inspired Neural Architecture Search for Few-Shot Learningcited by 22
- 2021No-Regret Dynamics in the Fenchel Game: A Unified Framework for Algorithmic Convex Optimizationcited by 20
- 2022Meta-Gradients in Non-Stationary Environmentscited by 11
- 2016Learning to Learn without Gradient Descent by Gradient Descentcited by 9
- 2018Learning to Explore with Meta-Policy Gradientcited by 9
Built on by
Citing papers per year: 2023: 2.
2 papers.