Meta-learning

The study of systems that use experience across tasks to learn how to adapt to a new task.

The study of systems that use experience across tasks to learn how to adapt to a new task.

is aResearch area All entries Open in graph

The object learned across tasks may be an initialization, representation, optimizer, or adaptation rule.

Connections

uses

First-order ANIL

Related publications

addressed by

First-order ANIL provably learns representations despite overparametrization

2024