Foundation Policies with Hilbert Representations

Overview

This is the official implementation of Hilbert Foundation Policies (HILPs).

This repository contains two different implementations of HILPs:

• For the zero-shot RL experiments, please refer to the hilp_zsrl directory.
• For the goal-conditioned RL experiments, please refer to the hilp_gcrl directory.

Visit our project page for more details.

License

MIT