meta-learning
Meta-learning, often referred to as 'learning to learn', involves training models in such a way that they can quickly adapt to new tasks using only a few examples. It aims to improve the efficiency and adaptability of learning algorithms.
- DataRater: Meta-Learned Dataset Curation
- Differentially Private Bilevel Optimization: Efficient Algorithms with Near-Optimal Rates
- Estimating Interventional Distributions with Uncertain Causal Graphs through Meta-Learning
- FSEO: Few-Shot Evolutionary Optimization via Meta-Learning for Expensive Multi-Objective Optimization
- Fast Rate Bounds for Multi-Task and Meta-Learning with Different Sample Sizes
- From Softmax to Score: Transformers Can Effectively Implement In-Context Denoising Steps
- Learning to Generalize: An Information Perspective on Neural Processes
- Learning to Learn with Contrastive Meta-Objective
- Learning to Learn with Contrastive Meta-Objective
- MaNGO — Adaptable Graph Network Simulators via Meta-Learning
- Meta-learning how to Share Credit among Macro-Actions
- MetaBox-v2: A Unified Benchmark Platform for Meta-Black-Box Optimization
- MetaGS: A Meta-Learned Gaussian-Phong Model for Out-of-Distribution 3D Scene Relighting
- MoEMeta: Mixture-of-Experts Meta Learning for Few-Shot Relational Learning
- NAVIX: Scaling MiniGrid Environments with JAX
- On the Stability and Generalization of Meta-Learning: the Impact of Inner-Levels
- PointMAC: Meta-Learned Adaptation for Robust Test-Time Point Cloud Completion
- Provably Efficient Multi-Task Meta Bandit Learning via Shared Representations
- TRiCo: Triadic Game-Theoretic Co-Training for Robust Semi-Supervised Learning
- Uncertain Knowledge Graph Completion via Semi-Supervised Confidence Distribution Learning