A meta-learning method learns a neural-network distance-generating function for mirror descent, matching or beating preconditioned baselines on few-shot image classification while providing an O(1/epsilon^2) convergence rate.
On the convergence the- ory of gradient-based model-agnostic meta-learning algorithms,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Learnable Loss Geometries with Mirror Descent for Scalable and Convergent Meta-Learning
A meta-learning method learns a neural-network distance-generating function for mirror descent, matching or beating preconditioned baselines on few-shot image classification while providing an O(1/epsilon^2) convergence rate.