Recurrent World Models Facilitate Policy Evolution
Ha, David, Schmidhuber, Jürgen
A generative recurrent neural network is quickly trained in an unsupervised manner to model popular reinforcement learning environments through compressed spatio-temporal representations. The world model's extracted features are fed into compact and simple policies trained by evolution, achieving state of the art results in various environments. We also train our agent entirely inside of an environment generated by its own internal world model, and transfer this policy back into the actual environment. Interactive version of paper at https://worldmodels.github.io
Sep-4-2018
- Country:
- North America
- United States
- Washington > King County
- Seattle (0.04)
- New York > New York County
- New York City (0.04)
- Massachusetts > Middlesex County
- Cambridge (0.04)
- Washington > King County
- Canada > Quebec
- Montreal (0.04)
- United States
- Europe
- Greece (0.04)
- United Kingdom > England
- Oxfordshire > Oxford (0.04)
- Germany > North Rhine-Westphalia
- Upper Bavaria > Munich (0.04)
- Finland > Uusimaa
- Helsinki (0.04)
- Asia
- North America
- Genre:
- Research Report (0.64)
- Industry:
- Leisure & Entertainment
- Games > Computer Games (1.00)
- Sports (0.67)
- Health & Medicine > Therapeutic Area
- Neurology (0.68)
- Leisure & Entertainment
- Technology: