Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Task Agnostic Meta-Learning for Few-Shot Learning
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Add entropy-based task-agnostic regularization to MAML | MAML · Omniglot 20-way 1-shot | accuracy: 90.81 $\pm$ 0.5\% -> 95.62 $\pm$ 0.5\% | Helped |
| Add entropy-based task-agnostic regularization to MAML | MAML · Omniglot 20-way 5-shot | accuracy: 97.49 $\pm$ 0.15\% -> 98.64 $\pm$ 0.13\% | Helped |
| Add Theil inequality regularization to MAML | MAML · Mini-ImageNet 5-way 5-shot | accuracy: 63.11 $\pm$ 0.92\% -> 65.94 $\pm$ 0.9\% | Helped |
| Add entropy-based task-agnostic regularization to MAML | MAML · Mini-ImageNet 5-way 5-shot | accuracy: 63.11 $\pm$ 0.92\% -> 66.05 $\pm$ 0.85\% | Helped |
| Add Theil inequality regularization to MAML | MAML · Mini-ImageNet 20-way 5-shot | accuracy: 19.29 $\pm$ 0.29\% -> 25.77 $\pm$ 0.33\% | Helped |
| Add GE(0) inequality regularization to Meta-SGD | Meta-SGD · Mini-ImageNet 20-way 5-shot | accuracy: 28.92 $\pm$ 0.35\% -> 29.75$\pm$ \textbf{0.34\%} | Helped |
| Use Theil inequality regularization instead of entropy regularization (ablation) | MAML · Mini-ImageNet 5-way 1-shot | accuracy: 49.33 $\pm$ 1.8\% vs 49.18 $\pm$ 1.8\% | No clear effect |
| Use inequality-minimization TAML rather than entropy-based TAML (ablation) | MAML · Mini-ImageNet 5-way 5-shot | accuracy: 66.05 $\pm$ 0.85\% vs 65.94 $\pm$ 0.9\% | No clear effect |
Closest papers by meaning
- Probabilistic Model-Agnostic Meta-Learning (NeurIPS 2018 · similarity 0.92)
- Meta-Learning without Memorization (ICLR 2020 · similarity 0.91)
- Task-Robust Model-Agnostic Meta-Learning (NeurIPS 2020 · similarity 0.90)
- Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks (ICML 2017 · similarity 0.90)
- Learning to Balance: Bayesian Meta-Learning for Imbalanced and Out-of-distribution Tasks (ICLR 2020 · similarity 0.89)
- Zero-Shot Task Adaptation with Relevant Feature Information (AAAI 2024 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.