AI Hub
All papers
Reinforcement Learning

Playing Atari with Deep Reinforcement Learning

DeepMind·December 19, 2013

Volodymyr Mnih, Koray Kavukcuoglu, David Silver

View on arXiv

TL;DR

The Deep Q-Network learns to play Atari games from raw pixels, reaching human-level play on many titles.

Why it matters

Launched deep reinforcement learning and put DeepMind on the map.

Related terms