Balancing Computation Load and Representation Expressivity in Parallel Hybrid Neural Networks

Open in new window