Search Results for author:
Found 2 papers, 1 papers with code
Can Active Sampling Reduce Causal Confusion in Offline Reinforcement Learning?
To answer this question, we consider a set of tailored offline reinforcement learning datasets that exhibit causal ambiguity and assess the ability of active sampling techniques to reduce causal confusion at evaluation.
Invariant Representations for Reinforcement Learning without Reconstruction
We study how representation learning can accelerate reinforcement learning from rich observations, such as images, without relying either on domain knowledge or pixel-reconstruction.
Cannot find the paper you are looking for? You can
Submit a new open access paper.
