Search Results for author:
Found 2 papers, 0 papers with code
Debiased Offline Representation Learning for Fast Online Adaptation in Non-stationary Dynamics
DORA incorporates an information bottleneck principle that maximizes mutual information between the dynamics encoding and the environmental data, while minimizing mutual information between the dynamics encoding and the actions of the behavior policy.
Programmatic Reinforcement Learning without Oracles
Our first contribution is a programmatically interpretable RL framework that conducts program architecture search on top of a continuous relaxation of the architecture space defined by programming language grammar rules.
Cannot find the paper you are looking for? You can
Submit a new open access paper.
