article

LDP: Learnable Dynamic Precision for Efficient Deep Neural Network Training and Inference
Max-Affine Spline Insights into Deep Network Pruning