Neuroevolution in deep neural networks: Current trends and future challenges

E Galván, P Mooney - IEEE Transactions on Artificial …, 2021 - ieeexplore.ieee.org
A variety of methods have been applied to the architectural configuration and learning or
training of artificial deep neural networks (DNN). These methods play a crucial role in the …

The child as hacker

JS Rule, JB Tenenbaum, ST Piantadosi - Trends in cognitive sciences, 2020 - cell.com
The scope of human learning and development poses a radical challenge for cognitive
science. We propose that developmental theories can address this challenge by adopting …

First return, then explore

A Ecoffet, J Huizinga, J Lehman, KO Stanley, J Clune - Nature, 2021 - nature.com
Reinforcement learning promises to solve complex sequential-decision problems
autonomously by specifying a high-level reward function only. However, reinforcement …

Bio-inspired computation: Where we stand and what's next

J Del Ser, E Osaba, D Molina, XS Yang… - Swarm and Evolutionary …, 2019 - Elsevier
In recent years, the research community has witnessed an explosion of literature dealing
with the mimicking of behavioral patterns and social phenomena observed in nature towards …

Aps: Active pretraining with successor features

H Liu, P Abbeel - International Conference on Machine …, 2021 - proceedings.mlr.press
We introduce a new unsupervised pretraining objective for reinforcement learning. During
the unsupervised reward-free pretraining phase, the agent maximizes mutual information …

Go-explore: a new approach for hard-exploration problems

A Ecoffet, J Huizinga, J Lehman, KO Stanley… - ar** elites
JB Mouret, J Clune - arxiv preprint arxiv:1504.04909, 2015 - arxiv.org
Many fields use search algorithms, which automatically explore a search space to find high-
performing solutions: chemists search through the space of molecules to discover new …

Paired open-ended trailblazer (poet): Endlessly generating increasingly complex and diverse learning environments and their solutions

R Wang, J Lehman, J Clune, KO Stanley - arxiv preprint arxiv:1901.01753, 2019 - arxiv.org
While the history of machine learning so far largely encompasses a series of problems
posed by researchers and algorithms that learn their solutions, an important question is …