• Publications
  • Influence
Beyond short snippets: Deep networks for video classification
TLDR
We propose and evaluate several deep neural network architectures to combine image information across a video over longer time periods than previously attempted. Expand
  • 1,561
  • 134
  • PDF
Deep Recurrent Q-Learning for Partially Observable MDPs
TLDR
We introduce the Deep Recurrent Q-Network (DRQN), a combination of a Long Short Term Memory (LSTM) (Hochreiter and Schmidhuber 1997) and a Deep Q- Network. Expand
  • 646
  • 99
  • PDF
Revisiting the Arcade Learning Environment: Evaluation Protocols and Open Problems for General Agents
TLDR
We introduce a new version of the ALE that supports multiple game modes and provides a form of stochasticity we call sticky actions. Expand
  • 195
  • 34
  • PDF
Leveraging Grammar and Reinforcement Learning for Neural Program Synthesis
TLDR
We introduce a training procedure that directly maximizes the probability of generating syntactically correct programs that fulfill the specification, which can be checked efficiently. Expand
  • 85
  • 20
  • PDF
Deep Reinforcement Learning in Parameterized Action Space
TLDR
We demonstrate reliable learning, from scratch, of RoboCup soccer policies capable of goal scoring. Expand
  • 166
  • 17
  • PDF
TextWorld: A Learning Environment for Text-based Games
TLDR
We introduce TextWorld, a sandbox learning environment for the training and evaluation of RL agents on text-based games. Expand
  • 77
  • 16
  • PDF
A Neuroevolution Approach to General Atari Game Playing
TLDR
This paper addresses the challenge of learning to play many different video games with little domain-specific knowledge. Expand
  • 140
  • 9
  • PDF
Autonomous Intersection Management: Multi-intersection optimization
TLDR
A multiagent systems approach to traffic intersection management called Autonomous Intersection Management (AIM) in which vehicles (driver agents) call ahead to a reservation manager agent at the intersection to reserve conflict-free trajectories in intersection space-time. Expand
  • 101
  • 8
  • PDF
HyperNEAT-GGP: a hyperNEAT-based atari general game player
TLDR
This paper considers the challenge of enabling agents to learn with as little domain-specific knowledge as possible. Expand
  • 55
  • 7
  • PDF
Counting to Explore and Generalize in Text-based Games
TLDR
We propose a recurrent RL agent with an episodic exploration mechanism that helps discovering good policies in text-based game environments. Expand
  • 26
  • 6
  • PDF