Adaptive Posterior Learning: few-shot learning with a surprise-based memory module.

Tiago Ramalho, Marta Garnelo

Browse the full ICLR paper archive.