research
∙
05/15/2023
MIMEx: Intrinsic Rewards from Masked Input Modeling
Exploring in environments with high-dimensional observations is hard. On...
share
research
∙
10/28/2021
Learning to Ground Multi-Agent Communication with Autoencoders
Communication requires having a common language, a lingua franca, betwee...
research
∙
04/28/2020
Visual Grounding of Learned Physical Models
Humans intuitively recognize objects' physical properties and predict th...
research
∙
09/15/2019
