Definition
The "learning to learn" paradigm, where an AI system is trained to efficiently adapt to and solve new, unseen tasks using knowledge gained from a distribution of previous tasks.
In short: The "learning to learn" paradigm, where an AI system is trained to efficiently adapt to and solve new, unseen tasks using knowledge gained from a distribution of previous tasks.