Meta-Learning of Structured Task Distributions in Humans and Machines

Sreejan Kumar
Sreejan Kumar
Ishita Dasgupta
Ishita Dasgupta

international conference on learning representations, 2021.

Cited by: 0|Views17
Weibo:
We developed a novel meta-learning task with a structured task distribution and statistically equivalent "null" task distribution to show humans are more adept at the former whereas current meta-learning agents are more adept at the latter.

Abstract:

In recent years, meta-learning, in which a model is trained on a family of tasks (i.e. a task distribution), has emerged as an approach to training neural networks to perform tasks that were previously assumed to require structured representations, making strides toward closing the gap between humans and machines. However, we argue that e...More
0
Full Text
Bibtex
Weibo
Your rating :
0

 

Tags
Comments