Learning General Optimal Policies with Graph Neural Networks: Expressive Power, Transparency, and Limits

Open in new window