Implicit Under-Parameterization Inhibits Data-Efficient Deep Reinforcement Learning

The dataset used in the paper is a collection of expert demonstrations for various tasks, including robotic manipulation, maze navigation, and Atari games.

BibTex: