Automatic Data Augmentation for Generalization in Reinforcement Learning

By Roberta Raileanu, Maxwell Goldstein, Denis Yarats, Ilya Kostrikov, Rob Fergus. 2021

139 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

1. Khimya Khetarpal
2. Matthew Riemer
3. Tim Rocktaschel
4. Irina Rish
5. Edward Grefenstette
6. Sergey Levine
7. Doina Precup
8. Yoshua Bengio

Citations in the corpus

• Towards Continual Reinforcement Learning: A Review and Perspectives (2020, background): “For example, agents have in some settings learned to generate their own imagined environments or scenarios for future play (Wang et al., 2019; Raileanu et al., 2021; Chen et al., 2021; Lee and Chung, 2021), their own goals (Florensa et al.”
• Prioritized Level Replay (2020, methodology): “1 ) leads to the largest statistically significant gains in mean normalized test and train returns and reduction in generalization gap compared to uniform sampling, outperforming all other methods besides UCB-DrAC + PLR. PLR combined with UCB-DrAC sees the most drastic improvements in these metrics.”