Few-Shot Learning: Prototypical Networks, MAML, and In-Context Adaptation

Status: public · Confidence: medium (0.78) · Basis: verified_sources

## TL;DR
Few-Shot Learning: Prototypical Networks, MAML, and In-Context Adaptation: Few-shot learning aims to learn new classes or tasks from only a small number of labeled examples.

## Core Explanation
Important approaches include metric learning, prototype-based classification, and meta-learning. These methods are evaluated with episodes that mimic learning from limited support examples.

## Further Reading

- [Matching Networks for One Shot Learning](https://arxiv.org/abs/1606.04080)
- [Prototypical Networks for Few-shot Learning](https://arxiv.org/abs/1703.05175)
- [Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks](https://arxiv.org/abs/1703.03400)