RECORD
On Enforcing Better Conditioned Meta-Learning for Rapid Few-Shot Adaptation
Machine Learning · Submitted 15 Jun 2022
abstract compiled · 2026-09-23 · no language-model review
The authors' main claim Our comprehensive evaluations show that the proposed method significantly outperforms its unconstrained counterpart especially during initial adaptation steps, while achieving comparable or better overall results on several few-shot classification tasks -- creating the possibility of dynamically choosing the number of adaptation steps at inference time.
The compiled record
The paper's own sentences, arranged. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.
Open the compiled record
The abstract, sentence by sentence
Inspired by the concept of preconditioning, we propose a novel method to increase adaptation speed for gradient-based meta-learning methods without incurring extra parameters. We demonstrate that recasting the optimization problem to a non-linear least-squares formulation provides a principled way to actively enforce a $\textit{well-conditioned}$ parameter space for meta-learning models based on the concepts of the condition number and local curvature. Our comprehensive evaluations show that the proposed method significantly outperforms its unconstrained counterpart especially during initial adaptation steps, while achieving comparable or better overall results on several few-shot classification tasks -- creating the possibility of dynamically choosing the number of adaptation steps at inference time.
background contribution method result interpretation main claim
| # | sentence | role | the authors' own work | states the main conclusion | strength |
|---|---|---|---|---|---|
| s0 | Inspired by the concept of preconditioning, we propose a novel method to increase adaptation speed for gradient-based meta-learning methods without incurring extra parameters. | contribution 1.00 | 0.77 | 0.06 | asserted 0.92 |
| s1 | We demonstrate that recasting the optimization problem to a non-linear least-squares formulation provides a principled way to actively enforce a $\textit{well-conditioned}$ parameter space for meta-learning models based on the concepts of the condition number and local curvature. | method 0.84 | 0.85 | 0.12 | asserted 0.99 |
| s2 | Our comprehensive evaluations show that the proposed method significantly outperforms its unconstrained counterpart especially during initial adaptation steps, while achieving comparable or better overall results on several few-shot classification tasks -- creating the possibility of dynamically choosing the number of adaptation steps at inference time. | result 0.97 | 0.96 | 0.82 | asserted 0.94 |
What was examined, and what was not
- arXiv listing abstract v2
- full text
- figures and tables
- references and citing sentences
- relations to other papers
- P abstract
- S sections
- T body sentences
- C citations
- R cross-paper relations
The paper
Original abstract
Inspired by the concept of preconditioning, we propose a novel method to increase adaptation speed for gradient-based meta-learning methods without incurring extra parameters. We demonstrate that recasting the optimization problem to a non-linear least-squares formulation provides a principled way to actively enforce a $\textit{well-conditioned}$ parameter space for meta-learning models based on the concepts of the condition number and local curvature. Our comprehensive evaluations show that the proposed method significantly outperforms its unconstrained counterpart especially during initial adaptation steps, while achieving comparable or better overall results on several few-shot classification tasks -- creating the possibility of dynamically choosing the number of adaptation steps at inference time.
Human review
How the literature uses it
What this paper does with the arXiv papers it cites
What it cites · 2 sentences without a confident label
Cites as background 2 papers
-
Gradient-based meta-learning as introduced in MAML [ref] has proved to be a versatile tool to successfully address such few-shot problems, and has inspired a significant number of follow-up works tackling a diverse set of applications like regression [ref] , image classification [ref] , reinforcement [ref] and online continual learning [ref] .
On First-Order Meta-Learning Algorithms · Introduction · 0.98
-
Early methods use LSTMs to update models iteratively by learning both the initialisation and update strategy via meta-learning [ref] but are difficult to train [ref] .
Meta-SGD: Learning to Learn Quickly for Few-Shot Learning · Related work · 0.91
Discussion
Know something about this paper? Leave a comment, or highlight any passage above to comment on it.