• Psychology, Computer Science, Mathematics
  • Published in CogSci 2016

Learning to reinforcement learn

@article{Wang2016LearningTR,
  title={Learning to reinforcement learn},
  author={Jane X. Wang and Zeb Kurth-Nelson and Hubert Soyer and Joel Z. Leibo and Dhruva Tirumala and R{\'e}mi Munos and Charles Blundell and Dharshan Kumaran and Matt M. Botvinick},
  journal={ArXiv},
  year={2016},
  volume={abs/1611.05763}
}
In recent years deep reinforcement learning (RL) systems have attained superhuman performance in a number of challenging task domains. However, a major limitation of such applications is their demand for massive amounts of training data. A critical present objective is thus to develop deep RL methods that can adapt rapidly to new tasks. In the present work we introduce a novel approach to this challenge, which we refer to as deep meta-reinforcement learning. Previous work has shown that… CONTINUE READING

Citations

Publications citing this paper.
SHOWING 1-10 OF 244 CITATIONS

Evaluation of Pretraining Methods for Deep Reinforcement Learning

VIEW 4 EXCERPTS
CITES BACKGROUND
HIGHLY INFLUENCED

Meta Reinforcement Learning with Autonomous Inference of Subtask Dependencies

VIEW 5 EXCERPTS
CITES METHODS & BACKGROUND
HIGHLY INFLUENCED

Catecholaminergic modulation of meta-learning

VIEW 10 EXCERPTS
CITES BACKGROUND
HIGHLY INFLUENCED

Learning to reinforcement learn for Neural Architecture Search

VIEW 10 EXCERPTS
CITES BACKGROUND & METHODS
HIGHLY INFLUENCED

Meta-Learning with Warped Gradient Descent

VIEW 10 EXCERPTS
CITES METHODS & BACKGROUND
HIGHLY INFLUENCED

FILTER CITATIONS BY YEAR

2016
2020

CITATION STATISTICS

  • 37 Highly Influenced Citations

  • Averaged 76 Citations per year from 2017 through 2019

  • 30% Increase in citations per year in 2019 over 2018

References

Publications referenced by this paper.
SHOWING 1-10 OF 53 REFERENCES

Meta-reinforcement learning: a bridge between prefrontal and dopaminergic function

  • Jane X Wang, Zeb Kurth-Nelson, +4 authors Matthew Botvinick
  • In Cosyne Abstracts,
  • 2017
VIEW 4 EXCERPTS
HIGHLY INFLUENTIAL

Finite-time Analysis of the Multiarmed Bandit Problem

VIEW 9 EXCERPTS
HIGHLY INFLUENTIAL

The formation of learning sets.

VIEW 5 EXCERPTS
HIGHLY INFLUENTIAL

Reinforcement Learning with Unsupervised Auxiliary Tasks

VIEW 4 EXCERPTS

Human-level control through deep reinforcement learning

VIEW 4 EXCERPTS

Adaptive behavior with fixed weights in RNN: an overview

VIEW 4 EXCERPTS
HIGHLY INFLUENTIAL